阅读说明：本技术 用于监测胎儿健康的系统和方法 (System and method for monitoring fetal health ) 是由 J·彭德斯 E·迪 M·阿尔蒂尼 于 2018-05-14 设计创作，主要内容包括：一种用于在怀孕期间随时间推移监测胎儿健康的系统包含：与孕妇耦合的传感器；与所述传感器通信耦合的处理器；以及其上存储有非暂时性处理器可执行指令的计算机可读介质。所述指令的执行使所述处理器执行包含以下的方法：从传感器获取信号；处理所述信号以从所述信号中识别并提取感兴趣参数；以及分析所述感兴趣参数以确定胎儿健康程度。所述感兴趣参数可以包含以下中的一个或多个：平均胎儿心率、平均胎儿心率变异性、胎儿踢动或运动计数、平均胎盘氧合水平、平均胎盘温度、平均胎盘pH、平均羊水量、胎儿心率曲线、胎儿心率变异性曲线和胎动曲线。(A system for monitoring fetal health over time during pregnancy comprising: a sensor coupled to the pregnant woman; a processor communicatively coupled with the sensor; and a computer readable medium having stored thereon non-transitory processor-executable instructions. Execution of the instructions causes the processor to perform a method comprising: acquiring a signal from a sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health. The parameter of interest may comprise one or more of: average fetal heart rate, average fetal heart rate variability, fetal kick or movement count, average placental oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, a fetal heart rate curve, a fetal heart rate variability curve, and a fetal movement curve.)

1. A system for monitoring fetal health over time during pregnancy, the system comprising:

a sensor coupled to the pregnant woman;

a processor communicatively coupled with the sensor; and

a computer-readable medium having stored thereon non-transitory processor-executable instructions, wherein execution of the instructions causes the processor to perform a method comprising:

acquiring a signal from the sensor;

processing the signal to identify and extract a parameter of interest from the signal; and

analyzing the parameter of interest to determine fetal health.

2. The system of claim 1, wherein the method performed by the processor further comprises comparing the parameter of interest to a fetal health index.

3. The system of claim 1, wherein the method performed by the processor further comprises tracking the parameter of interest over time to develop personalized fetal health trends.

4. The system of claim 3, wherein the method performed by the processor further comprises:

identifying deviations from the personalized fetal health trend; and

analyzing the deviation to determine whether the deviation is indicative of a change in fetal health.

5. The system of claim 1, wherein the method performed by the processor further comprises:

tracking the parameter of interest over time;

identifying deviations from population-level fetal health trends; and

analyzing the deviation to determine whether the deviation is indicative of a change in fetal health.

6. The system of claim 4 or 5, wherein analyzing the deviation is performed by one of thresholding, a machine learning algorithm, and regression modeling.

7. The system of claim 6, wherein the machine learning algorithm comprises one or more of a generalized linear model, a support vector machine, and a random forest.

8. The system of claim 5, wherein the population-level fetal health trend is derived from community data in a database.

9. The system of claim 8, wherein the community data comprises recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

10. The system of claim 1, wherein the system comprises a plurality of sensors.

11. The system of claim 10, wherein acquiring a signal comprises acquiring a plurality of signals.

12. The system of claim 1, wherein a plurality of parameters are extracted.

13. The system of claim 12, wherein the plurality of parameters include physiological, activity and behavior parameters.

14. The system of claim 1, wherein the sensors comprise one or more sensors configured to measure one or more of: fetal movement, fetal cardiac electrical activity, fetal heart sounds, fetal heart rate variability, fetal oxygenation, amniotic fluid volume, placental oxygenation, placental temperature, placental pH, fetal respiration, fetal location, fetal orientation, and fetal distress.

15. The system of claim 1, wherein the sensor senses one or more of biopotential signals, inertial signals, acoustic signals, ultrasonic signals, bioimpedance signals, optical signals, near infrared spectroscopy signals, electrochemical signals, and temperature signals.

16. The system of claim 1, wherein the parameter of interest comprises one or more of: average fetal heart rate, average fetal heart rate variability, average fetal heart beat, fetal kick count, fetal movement count, fetal oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, a fetal heart rate curve, a fetal heart rate variability curve, and a fetal movement curve.

17. The system of claim 1, further comprising a portable wearable sensor patch, the sensor patch comprising the sensor, the processor, and the computer-readable medium.

18. The system of claim 17, wherein the wearable sensor patch further comprises a wireless antenna for communicating with a mobile computing device.

19. The system of claim 1, wherein the sensor is located on or in a portable wearable sensor patch, the sensor patch further comprising electronic circuitry and a wireless antenna, and wherein the sensor patch is in wireless communication with a mobile computing device comprising the processor and the computer-readable medium.

20. The system of claim 1, wherein the method performed by the processor further comprises one or more of: generating an alert, providing feedback to the pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

21. The system of claim 1, wherein the method performed by the processor further comprises notifying a healthcare provider of the fetal health level.

22. The system of claim 1, wherein the method performed by the processor further comprises determining a probability of fetal distress.

23. The system of claim 22, wherein the method performed by the processor further comprises determining a degree of certainty that is consistent with the determined probability.

24. The system of claim 1, wherein the method performed by the processor further comprises determining a probability that a fetus is healthy.

25. The system of claim 24, wherein the method performed by the processor further comprises determining a degree of certainty that is consistent with the determined probability.

26. The system of claim 1, wherein analyzing the parameter of interest further comprises:

comparing the parameter of interest to a threshold.

27. The system of claim 26, wherein the probability of fetal health is higher if the parameter of interest is above the threshold.

28. The system of claim 26, wherein the probability of fetal distress is higher if the parameter of interest is below the threshold.

29. The system of claim 1, wherein the step of analyzing the parameter of interest further comprises:

analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

30. A computer-implemented method for longitudinally monitoring fetal health during pregnancy outside of a hospital environment, the method comprising:

acquiring a signal from a sensor;

processing the signal to identify and extract a parameter of interest from the signal; and

analyzing the parameter of interest to determine fetal health.

31. The method of claim 30, further comprising comparing the extracted parameter of interest to a fetal health index.

32. The method of claim 30, further comprising tracking the parameter of interest over time to develop personalized fetal health trends.

33. The method of claim 32, further comprising:

identifying deviations from the personalized fetal health trend; and

analyzing the deviation to determine whether the deviation is indicative of a change in fetal health.

34. The method of claim 30, further comprising:

tracking the parameter of interest over time;

identifying deviations from population-level fetal health trends; and

analyzing the deviation to determine whether the deviation is indicative of a change in fetal health.

35. The method of claim 33 or 34, wherein analyzing the deviation is performed by a machine learning algorithm.

36. The method of claim 35, wherein the machine learning algorithm comprises one or more of a generalized linear model, a support vector machine, and a random forest.

37. The method of claim 34, wherein the population-level fetal health trend is derived from community data in a database.

38. The method of claim 37, wherein the community data comprises recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

39. The method of claim 30, further comprising acquiring a plurality of signals.

40. The method of claim 39, further comprising extracting a plurality of parameters of interest.

41. The method of claim 30, wherein the sensor senses one or more of biopotential signals, inertial signals, acoustic signals, ultrasonic signals, bioimpedance signals, optical signals, near infrared spectroscopy signals, electrochemical signals, and temperature signals.

42. The method of claim 30, wherein the parameter of interest comprises one or more of: average fetal heart rate, average fetal heart rate variability, average fetal heart beat, fetal kick count, fetal movement count, fetal oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, a fetal heart rate curve, a fetal heart rate variability curve, and a fetal movement curve.

43. The method of claim 30, further comprising one or more of: generating an alert, providing feedback to a pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

44. The method of claim 30, further comprising notifying a healthcare provider of the fetal health level.

45. The method of claim 30, further comprising determining a probability of fetal distress.

46. The method of claim 45, further comprising determining a degree of certainty that is consistent with a determined probability.

47. The method of claim 30, further comprising determining a probability that a fetus is healthy.

48. The method of claim 47, further comprising determining a degree of certainty that is consistent with a determined probability.

49. The method of claim 30, wherein the step of analyzing the parameter of interest further comprises:

comparing the parameter of interest to a threshold.

50. The method of claim 49, wherein the probability of fetal health is higher if the parameter of interest is above the threshold.

51. The method of claim 49, wherein the probability of fetal distress is higher if the parameter of interest is below the threshold.

52. The method of claim 30, further comprising analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

53. The method of claim 30, further comprising:

tracking a plurality of the parameters of interest over time at a population level; and

a fetal health index is formed from the tracked parameter of interest.

54. The method of claim 53, further comprising:

comparing the extracted parameter of interest to the fetal health index to determine a fetal health level.

Technical Field

The present invention relates generally to the field of fetal health, and more particularly to a new and useful system and method for monitoring fetal health.

Background

The ability of the expectant mother to accurately monitor the health of the infant during pregnancy is critical to the early detection and potential treatment of adverse conditions. To date, there has not been any device or system that allows the expectant mother to track the health of their baby over time while performing daily activities or staying in their comfortable home during pregnancy (health/wellbanding). Such devices and/or systems are only available to clinical professionals in hospitals. Further, the present systems and devices are configured to provide information to a physician, nurse, doctor or other healthcare provider that can read and/or analyze the information and provide recommendations to the expectant mother. However, lay people (e.g., without clinical or medical training) often cannot understand this information itself.

Currently, hospitals track fetal health during discrete time points using fetal movement measurements, fetal heart rate (fHR) measurements, or more comprehensively using BioPhysical Profile (BPP). BPP combines fHR monitoring with fetal ultrasound to assess fetal health. Since ultrasound is used, BPP cannot be continuously performed. During BPP, the heart rate, respiration, exercise, muscle tone and amniotic fluid levels of the infants were analyzed and scored by physicians. Typically, the expectant mother is advised to perform BPP after 32 weeks of this marking or in some cases after 24 weeks. Further, BPP is generally only available to expectant mothers at high risk for pregnancy.

Another fetal health metric currently in use is kick count. The physician asks the expectant mother to track fetal movements at home. However, only about one third of all kicks are actually detected, resulting in a high error rate. Further, due to fear of missing a kick or movement, the mother-to-be is prohibited from doing anything other than tracking the fetal movement.

As mentioned above, monitoring fetal health is critical to modern obstetrics. Although fetal movement and fHR are routinely used as indicators of fetal health, accurate non-invasive long-term monitoring of fetal movement and fHR remains challenging. To reduce risk, accelerometer-based systems have been developed to address common problems in ultrasound motion, and ECG-based systems have been developed to address common problems in heart rate monitoring. These systems enable monitoring of fetal movement during pregnancy. However, many of these self-administered wearable sensors lack optimal settings and lack signal processing and machine learning techniques for detecting fetal movement and fHR.

Accordingly, there is a need for systems and methods that allow a prospective mother to receive understandable data about their developing baby and monitor the health of the fetus over time. The present invention provides such a new and useful system and method.

Disclosure of Invention

The following is a non-exhaustive list of certain aspects of the present technology. These and other aspects are described in the following disclosure.

Some aspects include a system for monitoring fetal health over time during pregnancy, the system comprising: a sensor coupled to the pregnant woman; a processor communicatively coupled with the sensor; and a computer readable medium having stored thereon non-transitory processor-executable instructions, wherein execution of the instructions causes the processor to perform a method comprising: acquiring a signal from the sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health.

One aspect of the present disclosure relates to a system for monitoring fetal health over time during pregnancy. In some embodiments, a system comprises: a sensor coupled to the pregnant woman; a processor communicatively coupled with the sensor; and a computer readable medium having stored thereon non-transitory processor-executable instructions. In some embodiments, execution of the instructions causes the processor to perform a method comprising: acquiring a signal from the sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health.

In some embodiments, the method performed by the processor further includes comparing the parameter of interest to a fetal health index.

In some embodiments, the method performed by the processor further includes tracking the parameter of interest over time to develop personalized fetal health trends.

In some embodiments, the method performed by the processor further includes: identifying deviations from the personalized fetal health trend; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

In some embodiments, the method performed by the processor further includes: tracking the parameter of interest over time; identifying deviations from population-level fetal health trends; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress. In some embodiments, analyzing the deviation is performed by one of thresholding, a machine learning algorithm, and regression modeling.

In some embodiments, the machine learning algorithm includes one or more of a generalized linear model, a support vector machine, and a random forest.

In some embodiments, the population-level fetal health trend is derived from community data in a database. In some embodiments, the community data includes recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

In some embodiments, the system includes a plurality of sensors.

In some embodiments, acquiring the signal includes acquiring a plurality of signals.

In some embodiments, a plurality of parameters are extracted. In some embodiments, the plurality of parameters includes physiological, activity and behavioral parameters.

In some embodiments, the sensors include one or more sensors configured to measure one or more of: fetal movement, fetal cardiac electrical activity, fetal heart sounds, fHR, fetal heart rate variability (fHRV), fetal oxygenation level, amniotic fluid volume, placental oxygenation, placental temperature, placental pH, fetal respiration, fetal position, fetal orientation, and fetal distress. In some embodiments, the sensor senses one or more of biopotential signals, inertial signals, acoustic signals, ultrasonic signals, bioimpedance signals, optical signals, near infrared spectroscopy signals, electrochemical signals, and temperature signals.

In some embodiments, the parameter of interest comprises one or more of an average fHR, an average fHRV, an average fetal heart beat, a fetal kick count, a fetal movement count, a fetal oxygenation level, an average placental temperature, an average placental pH, an average amniotic fluid volume, a fHR curve, a fHRV curve, and a fetal movement curve.

In some embodiments, the system further includes a portable wearable sensor patch including the sensor, the processor, and the computer-readable medium. In some embodiments, the wearable sensor patch further includes a wireless antenna for communicating with a mobile computing device.

In some embodiments, the sensor is located on or in a portable wearable sensor patch, the sensor patch further comprising electronic circuitry and a wireless antenna, and wherein the sensor patch is in wireless communication with a mobile computing device comprising the processor and the computer-readable medium.

In some embodiments, the method performed by the processor further includes one or more of: generating an alert, providing feedback to a pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

In some embodiments, the method performed by the processor further includes notifying a healthcare provider of the fetal health level.

In some embodiments, the method performed by the processor further comprises determining a probability of fetal distress. In some embodiments, the method performed by the processor further includes determining a degree of certainty that is consistent with the determined probability. In some embodiments, the method performed by the processor further includes determining a probability that the fetus is healthy.

In some embodiments, analyzing the parameter of interest further comprises: comparing the parameter of interest to a threshold.

In some embodiments, the probability of fetal health is higher if the parameter of interest is above the threshold. In some embodiments, the probability of fetal distress is higher if the parameter of interest is below the threshold.

In some embodiments, analyzing the parameter of interest further comprises: analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

Another aspect of the present disclosure relates to a computer-implemented method of longitudinally monitoring fetal health during pregnancy outside of a hospital environment. In some embodiments, the method comprises: acquiring a signal from a sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health.

In some embodiments, the method further comprises comparing the extracted parameter of interest to a fetal health index.

In some embodiments, the method further includes tracking the parameter of interest over time to develop personalized fetal health trends. In some embodiments, the method further comprises: identifying deviations from the personalized fetal health trend; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

In some embodiments, the method further comprises: tracking the parameter of interest over time; identifying deviations from population-level fetal health trends; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

In some embodiments, analyzing the deviation is performed by a machine learning algorithm.

In some embodiments, the machine learning algorithm comprises one or more of a generalized linear model, a support vector machine, and a random forest.

In some embodiments, the population-level fetal health trend is derived from community data in a database. In some embodiments, the community data includes recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

In some embodiments, the method further comprises acquiring a plurality of signals.

In some embodiments, the method further comprises extracting a plurality of parameters of interest.

In some embodiments, the sensor senses one or more of biopotential signals, inertial signals, acoustic signals, ultrasonic signals, bioimpedance signals, optical signals, near infrared spectroscopy signals, electrochemical signals, and temperature signals.

In some embodiments, the parameter of interest comprises one or more of an average fHR, an average fHRV, an average fetal heart beat, a fetal kick count, a fetal movement count, an average placental oxygenation level, an average placental temperature, an average placental pH, an average amniotic fluid volume, a fHR curve, a fHRV curve, and a fetal movement curve.

In some embodiments, the method further comprises one or more of: generating an alert, providing feedback to a pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

In some embodiments, the method further comprises notifying a healthcare provider of the fetal health level.

In some embodiments, the method further comprises determining a probability of fetal distress. In some embodiments, the method further includes determining a degree of certainty that is consistent with the determined probability. In some embodiments, the method further comprises determining a probability of fetal health.

In some embodiments, analyzing the parameter of interest further comprises: comparing the parameter of interest to a threshold.

In some embodiments, the probability of fetal health is higher if the parameter of interest is above the threshold. In some embodiments, the probability of fetal distress is higher if the parameter of interest is below the threshold.

In some embodiments, the method further comprises analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

In some embodiments, the method further comprises: tracking a plurality of the parameters of interest over time at a population level; and forming a fetal health index based on the tracked parameter of interest.

In some embodiments, the method further comprises: comparing the extracted parameter of interest to the fetal health index to determine a fetal health level.

Drawings

The foregoing and other aspects of the present technology will be better understood when the present application is read in light of the following drawings in which like characters represent similar or identical elements:

fig. 1 depicts a block diagram of one embodiment of a system for determining fetal health of a pregnant female.

Fig. 2 depicts a block diagram of another embodiment of a system for determining fetal health of a pregnant female.

Fig. 3 depicts a block diagram of another embodiment of a system for determining fetal health of a pregnant female.

Fig. 4 depicts a top view of one embodiment of a sensor module that forms part of a system for determining fetal health of a pregnant female.

Fig. 5 depicts a top view of another embodiment of a sensor module that forms part of a system for determining fetal health of a pregnant female.

Fig. 6 depicts a perspective view of one embodiment of the sensor module applied to the abdominal region of a pregnant woman.

Fig. 7 depicts a perspective view of another embodiment of the sensor module applied to the abdominal region of a pregnant woman.

Fig. 8 depicts a perspective view of another embodiment of the sensor module applied to the abdominal region of a pregnant woman.

Fig. 9 depicts a flowchart of one embodiment of a computer-implemented method for longitudinally monitoring fetal health during pregnancy.

Fig. 10 depicts a flowchart of one embodiment of a computer-implemented method for longitudinally monitoring fetal health during pregnancy.

Fig. 11 depicts a flowchart of one embodiment of a computer-implemented method for longitudinally monitoring fetal health during pregnancy.

Fig. 12 depicts a perspective view of an embodiment of a fetal health sensor module that forms part of a system for determining fetal health of a pregnant female.

While the technology is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. The drawings may not be to scale. It should be understood, however, that the drawings and detailed description thereto are not intended to limit the inventive technique to the particular form disclosed, but on the contrary, the intention is to cover all modifications, equivalents and alternatives falling within the spirit and scope of the inventive technique as defined by the appended claims.

Detailed Description

To alleviate the problems described herein, the inventors must both invent solutions and, in some cases, equally importantly, reorganize the problem that is ignored (or not foreseen) by others in the field of fetal health. Indeed, the inventors wish to emphasize the difficulty of identifying those issues that are new and will become more apparent in the future if industry trends continue as the inventors expect. Further, because a number of problems are addressed, it should be understood that some embodiments are problem-specific and that not all embodiments address or provide each of the benefits of the conventional systems described herein. That is, various alternative refinements to address these issues are described below.

Monitoring fetal health during pregnancy is the most important and complex task of modern obstetrics. Since birth results are closely related to the development of fetal condition during pregnancy, several techniques have been proposed to monitor fetal health (e.g., exercise, fHR, etc.). Some methods require hospitalized or trained personnel, such as ultrasound, to rely on high frequency sound waves for generating images of the fetus, which can only be used for a limited time due to safety considerations.

Other methods for providing fetal activity monitoring or their equivalents such as continuous fetal heart monitoring (cardio monitoring) require cumbersome infrastructure and hospital visits, involve trained personnel to set up the device and process the resulting information. Therefore, the ability of these methods to monitor fetal movements outside of sporadic spot checks in a hospital environment is one of the main reasons for concern about passive methods (such as accelerometer-based solutions) for home monitoring.

Many accelerometer-based solutions use a single accelerometer placed in the abdominal region. This involves different criteria such as the number of sensors used, the presence of a reference accelerometer placed outside the abdominal region, and data analysis. The sensitivity and pertinence of this technique is low. As a result, the detection rate was about 50%, which was considered insufficient by researchers.

Higher detection rates can be achieved by adding a reference accelerometer that monitors maternal movement attributes using accelerometers placed outside the abdominal region, thus separating fetal movement from maternal movement and providing more accurate detection. The location of the accelerometer is critical because its placement in the upper thoracic region may still detect fetal movement and provide unusable results, and therefore, the accelerometer should be outside of the abdominal and thoracic regions. To date, no study has been reported on the differences in motion detection performance with or without the inclusion of a reference accelerometer. The techniques are often used as post-processing signaling to discard data rather than just inform the classification process. These are all suggestive of a disclaimer of this method or any other subject matter, as the present technology may be used to augment various methods similar to this method.

The data analysis techniques used so far have focused mainly on feature extraction by means of time-domain (e.g., magnitude of acceleration vectors) and frequency-domain signal processing techniques. Recently, a set of features has been classified as binary problems using machine learning techniques such as support vector machines. While determining the best feature may be a necessary first step, thresholding a single feature may provide poor results. The combination of multiple features and machine learning methods has the potential to perform fetal movement detection more accurately. Challenges may be encountered when classifying motion using supervised learning approaches. Fetal movement occurs only for a small fraction of the time during the measurement session and therefore requires the use of appropriate methods such as down sampling most categories (i.e. no movement). It may be necessary to evaluate not only the subset of data that was pre-selected by the user, but also the entire data stream. Another design choice problem is the window size of the computational features, the choice of classifiers, the feature selection method, the performance metrics for the analysis system, and the reference system for validating the fetal movement detection algorithm.

As noted above, most studies rely on ultrasound as a reference for fetal movement. Although ultrasound is a clinical standard, there are limitations. In some cases, as the fetus grows, starting from about week 20 of pregnancy, it is difficult to fully display the fetus given the limited field of view of the ultrasound probe. This particular concern may not be a problem during a hospital examination, but moving and repositioning the probe while measuring small accelerations reflected on the pregnant abdomen is not practical and may easily introduce noise.

Some embodiments alleviate some or all of the problems discussed above as well as others discussed below and those that will be apparent to those of ordinary skill in the art in view of the pending disputes in this field. Some embodiments may incorporate analysis of algorithm performance and tradeoffs with respect to a reference for accurate sensor numbering, sensor positioning, and data analysis to effectively detect fetal movements.

Some embodiments alleviate some or all of the problems discussed above by generating algorithms that highlight differences in the model, for example, by reducing the difference in Positive Predictive Value (PPV) by reducing false positives in the data set when the reference accelerometer is present and both short and long time windows are used for feature computation.

Some embodiments alleviate some or all of the problems discussed above and highlight how the model performs at different levels by analyzing detected and actual fetal kicks, e.g., recording individual fetal kicks at the level and total fetal kicks across the entire data set. This may allow individual movements to be accurately identified at incremental intervals, for example, intervals of about 20 minutes.

Some embodiments alleviate some or all of the problems discussed above by aggregating three types of motion, such as low, medium, and high motion, and analyzing the results according to the model's ability to accurately identify overall motion levels with respect to individual motions.

Another aspect of monitoring fetal health is fHR and fHRV detection. The fHR and fHRV tests most commonly used today are not suitable for long-term monitoring. For example, doppler ultrasound and fetal scalp electrodes are the most commonly used methods. Doppler ultrasound, while non-invasive, releases energy into the body, which requires constant supervision, making it unsuitable for continuous monitoring of fHR and fHRV. In addition, doppler ultrasound measurements still require input from trained medical personnel to make informed clinical decisions. Further, for example, fetal scalp electrodes are highly traumatic and require rupturing of the membrane and are therefore only useful during labor. Since the needle electrodes are screwed into the scalp of the fetus to obtain a fetal electrocardiogram (fECG) signal, there is a risk of infection and tissue damage. In general, the widespread use of doppler ultrasound and fetal scalp electrodes is not ideal for continuously or nearly continuously monitoring fetal health.

Electrophysiological measurements of the abdomen of the parturient also contain fcecg and enable extraction of fHR. However, the signal-to-noise ratio (SNR) of the signal is reduced compared to using fetal scalp electrodes, while maternal ecg (mcecg) is the main disturbance. Methods that attempt to address the SNR problem of electrophysiological measurements have greatly improved fHR detection in non-flow environments, but due to the computational complexity, it is still not feasible to continuously monitor the fetus in a flow environment using these methods.

Some embodiments alleviate some or all of the problems discussed above by reducing overall computational complexity using discrete-time continuous wavelet transforms while increasing the R-peak detection quality of mcgs.

Some embodiments mitigate some or all of the problems discussed above by improving the R peak detection quality via segment selection (e.g., heart rate limit, R-R interval, previous segment), threshold determination (e.g., previous threshold, maximum in segment, SNR estimate), SNR estimation (e.g., R peak height, QRS-outer maximum, log2 of ratio), and peak detection (e.g., when the first peak is greater than the threshold, the highest peak is selected within 0.05 to 5 seconds).

Systems and methods for monitoring fetal health are disclosed herein. In some embodiments, monitoring fetal health comprises monitoring fetal movement, fHR and/or fHRV. The systems and methods described herein generally include a sensor module for monitoring fetal health continuously or over time. The monitoring results may be provided to the pregnant woman or mother-as-you-go, gynecologist, obstetrician, pediatrician, other physician, nurse practitioner, veterinarian, other healthcare provider, mentor, midwife, other labor specialist, spouse, companion, parent, sibling, other family member, friend, healthcare facility administrator, or any other individual for whom the pregnant woman wishes to share such information.

As used herein, "pregnant woman," "pregnant woman," or "expectant mother" may be used interchangeably. One skilled in the art will recognize that each of the embodiments described herein can be used by pregnant mammals regardless of species.

As used herein, "infant," "fetus," or "developing infant" may be used interchangeably. One skilled in the art will recognize that each of the embodiments described herein can be used to monitor the health of a fetus regardless of the species.

As used herein, "parameter of interest" refers to a pattern, feature, characteristic, component, aspect, element, or attribute extracted from a sensor signal related to fetal health. The parameters of interest may include average fHR, average fHRV, average fetal heart beat, fetal kick count, fetal movement count, average placental oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, fHR curve, fHRV curve, and fetal movement curve.

As used herein, "fetal health index" refers to a comprehensive measure of fetal health that profiles and/or ranks one or more particular fetal health observations and/or measurements. One skilled in the art will recognize that the fetal health index described herein may be derived from one or more observations or measurements. Further, it will also be appreciated that a high fetal health index score may indicate a high probability of a healthy or distressed fetus, and a low fetal health index score may indicate a high probability of a healthy or distressed fetus.

System for controlling a power supply

In some embodiments, the above-described features may be implemented in a system 10 as shown in FIG. 1. It should be emphasized, however, that not all embodiments comprise all of the above features, provide all of the above advantages, or mitigate, in part or in whole, all of the above problems, and this is not to be taken as a limitation on any other description herein. Rather, a variety of independently useful techniques with various engineering and cost tradeoffs are described, and some embodiments may implement some of those techniques while other embodiments do not implement others.

As shown in fig. 1, in various embodiments, a system 10 for monitoring fetal health may include at least one sensor 12 in electrical communication with a processor 14 and a computer-readable medium (i.e., memory) 16. FIG. 1 illustrates a functional block diagram, and it is to be understood that the various functional blocks of the depicted system 10 are not necessarily separate structural elements. For example, in some embodiments, processor 14 and memory 16 may be embodied in a single chip or in two or more chips.

The sensors 12 detect fetal events (e.g., fetal kick, fetal movement, etc.), physiological characteristics (e.g., heart rate, placental oxygenation level, etc.), and/or environmental changes (e.g., amniotic fluid volume) and provide corresponding outputs or signals. In some embodiments, the system 10 includes one sensor 12; in some embodiments, the system 10 includes a plurality of sensors 12. For example, the sensors 12 may include one or more sensors configured to measure: fetal movement, fetal cardiac electrical activity, fetal heart sounds, fHR, fHRV, amniotic fluid volume, placental oxygenation, placental temperature, placental pH, fetal respiration, fetal position, fetal orientation, and/or fetal distress.

The sensor 12 of various embodiments is configured for placement on an exterior surface of a female body. In some embodiments, the sensor 12 may be reusable; in other embodiments, the sensor 12 is disposable. In at least some embodiments, the sensor 12 is configured for placement over the abdomen or abdominal region of a pregnant woman. In some embodiments, the sensor 12 forms part of a sensor module. Various sensor module embodiments are described in more detail below with reference to fig. 2-8.

The sensors 12 may include biopotential sensors, inertial sensors, acoustic sensors, ultrasonic sensors, bioimpedance sensors, optical sensors, near infrared spectroscopy sensors, electrochemical sensors, and/or temperature sensors. The biopotential sensor interacts with ionic charge carriers and converts the ionic current into a current that is read by a processor. A biopotential sensor as described herein may comprise at least one measurement electrode and at least one reference electrode. In some configurations, there is one reference electrode and multiple measurement electrodes in the biopotential sensor. The biopotential sensors may measure the ECG, electroencephalogram (EEG), or Electromyogram (EMG) of the fetus or expectant mother.

Inertial sensors as described herein include one or more accelerometers, gyroscopes, and/or magnetometers to measure specific forces (i.e., g-forces or mass-specific forces), angular velocities, and/or magnetic fields around the body. For example, an inertial sensor or inertial sensors of the system may be used to measure fetal movement, fetal position, and/or fetal orientation.

As described herein, an acoustic sensor, such as an ultrasound sensor, uses acoustic waves that propagate through a portion of the abdomen of a pregnant woman (which may include a portion of the uterus and/or fetus) to measure a characteristic of the pregnant woman, the uterus, the placenta, the fetus, or any other characteristic of the fetus or a structure supporting the growth of the fetus. As the sound waves propagate through the abdomen, one or more characteristics of the waves change, such as velocity, amplitude, and the like. These changes are monitored by the sensor and output as a sensor signal.

The bio-impedance sensors as described herein use electrical current to measure various cardiac parameters, such as a fetus or mother-to-be. The cardiac parameter may comprise stroke volume, heart rate, cardiac output, heart rate variability, or any other parameter known to one of skill in the relevant art. In some embodiments, one or more bio-impedance sensors are used to measure the amount of amniotic fluid. For example, accumulation of amniotic fluid (i.e., polyhydramnios) or absence of amniotic fluid (i.e., oligohydramnios) may be detected by one or more bioimpedance sensors. One non-limiting example of a bioimpedance sensor includes an impedance plethysmography sensor.

An optical sensor as described herein illuminates one or more areas of the skin and measures changes in light absorption or reflection. For example, optical sensors may be used to measure the oxygen saturation of the placenta, blood flow to various organs or appendages, blood pressure, or pulse. One non-limiting example of an optical sensor includes a photoplethysmogram.

Near infrared spectroscopy sensors as described herein use near infrared light to illuminate one or more regions of the skin and measure changes in electromagnetic absorption at this particular frequency band. It can be used for non-invasive assessment of placental function, for example by measuring placental oxygenation, blood flow, sugar levels or pH.

Electrochemical sensors as described herein use electrochemical reactions to measure the concentration of particular ions and can be used to measure the acidity or pH of body fluids such as sweat or interstitial fluid.

A temperature sensor as described herein may be used to measure the average placenta temperature. Non-limiting examples of temperature sensors include thermistors and thermocouples.

Returning to fig. 1, the processor 14 of fig. 1 may be a general purpose microprocessor, Digital Signal Processor (DSP), Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), or other programmable logic device or other discrete computer-executable components designed to perform the functions described herein. A processor may also be formed from a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other suitable configuration.

In some embodiments, processor 14 is coupled to memory 16 via one or more buses, in order to read information from memory 16 and optionally write information to the memory. Memory 16 may be any suitable computer readable medium that stores computer readable instructions for execution by processor 14. For example, a computer-readable medium may comprise one or more of RAM, ROM, flash memory, EEPROM, a hard drive, a solid state drive, or any other suitable device. In some embodiments, the computer readable instructions comprise software stored in a non-transitory format. The software may be programmed into the memory 16 or downloaded as an application onto the memory 16. The software may contain instructions for running an operating system and/or one or more programs or applications. The program or application, when executed by the processor 14, may cause the processor 14 to perform a method of monitoring fetal health over time during pregnancy. Some such methods are described in more detail elsewhere herein.

As shown in fig. 2 and 3, system 10 may further include a sensor module 18 and a mobile computing device 20. In some embodiments, the system 10 also includes a server 30. In some embodiments, such as the embodiment of FIG. 2, the sensors 12, the processor 14, and the memory 16 are each located on or in a sensor module 18. The electronic circuitry 15 and the wireless antenna 13 may also be provided on or in the sensor module 18. In such embodiments, the signal related to fetal health: sensed by the sensor 12; amplified, filtered, digitized and/or otherwise processed by the electronic circuitry 15; and analyzed by processor 14. Execution of the instructions stored in memory 16 causes processor 14 on sensor module 18 to perform one or more of the methods of monitoring fetal health described elsewhere herein. The analyzed data may be transmitted via the antenna 13 to one or both of the mobile computing device 20 and the server 30 for visual or audio presentation to the user, for additional analysis, and/or for storage.

In other embodiments, such as the embodiment of fig. 3, the sensor 12 is located on or in the sensor module 18 along with the electronic circuitry 15 and wireless antenna 13, while the mobile computing device 20 houses a processor 14 that performs the method of monitoring fetal health during pregnancy and a memory 16 that stores instructions for performing the method. In such embodiments, signals related to fetal health are sensed by the sensor 12 and amplified, filtered, digitized, and/or otherwise processed by the electronic circuitry 15, and the processed signals are transmitted to the mobile computing device 20 via the antenna 13. As described elsewhere herein, processor 14 of mobile computing device 20 analyzes the processed signals and determines the fetal health level. The analyzed data may be saved, shared with the contact, or presented to the user via the mobile computing device 20. In some such embodiments, some or all of the analysis data may be transmitted from mobile computing device 20 to server 30 for storage.

In some embodiments, the electronic circuitry 15 includes operational amplifiers, low-pass, high-pass, or band-pass filters, analog-to-digital (AD) converters, and/or other signal processing circuit components configured to amplify, filter, digitize, and/or otherwise process the physiological signal. The electronic circuit 15 may additionally contain a power source or power storage device, such as a battery or capacitor, to provide power to other electronic components. For example, the electronic circuit 15 may comprise a rechargeable (e.g. lithium ion) or disposable (e.g. alkaline) battery.

In some embodiments, the antenna 13 includes one or both of a receiver and a transmitter. The receiver receives data through the communication network and demodulates the data received through the communication network. The transmitter prepares data according to one or more network standards and transmits the data over a communication network. In some embodiments, the transceiver antenna 13 functions as both a receiver and a transmitter for two-way wireless communication. In addition to or in lieu of antenna 13, in some embodiments, a data bus is provided within sensor module 18 such that data may be transmitted from or received by sensor module 18 via a wired connection.

In some embodiments, there is one-way or two-way communication between sensor module 18 and mobile computing device 20, sensor module 18 and server 30, and/or mobile computing device 20 and server 30. Sensor module 18, mobile computing device 20, and/or server 30 may communicate wirelessly using bluetooth, bluetooth low energy, near field communication, infrared, WLAN, Wi-Fi, CDMA, LTE, other cellular protocols, other radio frequencies, or another wireless protocol. Additionally or alternatively, sending or transmitting information between sensor module 18, mobile computing device 20, and server 30 may be via a wired connection, such as IEEE 1394, Thunderbolt (Thunderbolt), Lightning (Lightning), FireWire (FireWire), DVI, HDMI, Serial (Serial), Universal Serial Bus (Universal Serial Bus), Parallel (parallell), Ethernet (Ethernet), Coaxial (coax), VGA, or PS/2.

In some embodiments, the mobile computing device 20 is a computing device that is packaged in a chassis that includes a visual display (e.g., a thin film transistor Liquid Crystal Display (LCD), a switch-in-place LCD, a resistive touchscreen LCD, a capacitive touchscreen LCD, an organic Light Emitting Diode (LED), an Active Matrix Organic LED (AMOLED), a super AMOLED, a retinal display, a tactile touchscreen, or Gorilla Glass (Gorilla Glass)), an audio output (e.g., a speaker), a central processing unit (e.g., a processor or microprocessor), an internal storage device (e.g., a flash drive), n components (e.g., a dedicated chip and/or sensor), and n radios (e.g., WLAN, LTE, WiFi, bluetooth, GPS) with or without touch response capability. In some embodiments, the mobile computing device 20 is a mobile phone, a smartphone, a smartwatch, smart glasses, smart contact lenses, or other wearable computing device, a tablet computer, a portable computer, a netbook, a notebook computer, or any other type of mobile computing device. In some embodiments, the mobile computing device 20 may be a personal computer.

In some embodiments, the display of the mobile computing device 20 may present a user interface for manual data entry by the pregnant female or automated data entry (e.g., automated data synchronization) from one or more clinic or hospital records. The user interface may include a user profile specification such as conception date, date of pre-delivery, number of weeks of pregnancy (e.g., calculated from conception date and/or date of pre-delivery), starting weight, current weight, weight over time, fetal sonograms, and/or any other information. The profile information may be used in combination with additional data and/or parameters to determine a fetal health threshold or a personalized fetal health trend, which will be described in more detail elsewhere herein.

In some embodiments, the server 30 is a database server, an application server, an internet server, or other remote server. In some embodiments, the server 30 may store user profile data, historical user data, historical community data, algorithms, machine learning models, software updates, or other data. Server 30 may share such data with mobile computing device 20 or sensor module 18, and server 30 may receive newly acquired user data from sensor module 18 and/or mobile computing device 20.

Some non-limiting examples of sensor modules 18 are depicted in fig. 4-8 and 12. By comparing the sensor modules of fig. 4-8 and 12, it can be readily appreciated that sensor module 18 can take many different form factors. The sensor module 18 of various embodiments has many different shapes, sizes, colors, materials, and levels of body fit. The sensor module 18 may be connected to, embedded in, or form part of: patches 40, 42 (e.g., fig. 4-6), a belt, waistband or strap 44 (e.g., fig. 7) or carpet/covering 46 (e.g., fig. 8), T-shirt, pants, underwear, or other garment or wearable accessory.

Referring to fig. 4, the fetal health monitoring device comprises an electrode patch 40 and a sensor module 18 which advantageously combine to monitor fetal movement, fHR, fHRV, mHR and/or at least one uterine contraction signal. The electrode patch 40 and the sensor module 18 may be one part or may be made of two separate parts. The two separate parts may be provided with mechanical and electrical systems for attachment to each other, such as a trimming system or magnets. Other embodiments are described in the specification.

Fig. 5 illustrates another embodiment of a fetal health monitoring device. By comparing fig. 4 and 5, it will be readily appreciated that the electrode patches 40, 42 or sensor module 18 may take many different form factors.

In other words, the fetal health monitoring device may take many different shapes, sizes, colors, materials, and levels of body conformity. The device may or may not take the form of plaster. For example, the device may be integrated in an article of clothing, in the form of an article of clothing or textile, or may take the form of a belt worn around the abdomen. For the last three examples, the electrode patches 40, 42 may be an integral part of the piece of clothing, garment or belt, or may be attached to such a piece of clothing, garment or belt.

Fig. 6 shows another embodiment of a fetal health monitoring device in which the electrode patch 42 and sensor module 18 may be integrated and packaged into one single component forming a separate device. For example, the fetal health monitoring device of fig. 6 may have at least three electrodes, including one measuring electrode located on one end of the device, one reference electrode located on the other end of the device, and one biasing electrode in the middle. This arrangement enables measurement of a path of biopotential signals, motion signals, fHR signals and/or fHRV signals in the horizontal direction. In some embodiments, the device of fig. 6 may have four electrodes, two measurement electrodes at the two ends, one reference electrode in the middle of the device and one bias electrode between the measurement and reference electrodes. Advantageously, a variation of the device of fig. 6 (not shown) may have five electrodes, two measurement electrodes located at the two ends of the device, one reference electrode located in the middle of the device, the other measurement electrode located below the reference electrode at 90 degrees to the line between the first three electrodes and one bias electrode located between the measurement electrode and the reference electrode. This configuration enables the measurement of two biopotential, motion, fHR and/or fHRV signals, one along the horizontal and one along the vertical direction. In some embodiments, electrode patch 42 does not include a reference electrode; in contrast, the reference electrode is absent or located on the back of the pregnant female. In further embodiments, the device may be attached to the body using an adhesive layer. In another embodiment, the adhesive layer may be replaceable by the user. In another exemplary embodiment, the device may be attached to the body using a belt or textile that can maintain the device in contact with the body.

Fig. 7 shows an exemplary embodiment of a fetal health monitoring device 44, in which the electrode patch and sensor module may be integrated in a textile or clothing accessory. Examples of a garment accessory may include, but are not limited to, a shirt, a T-shirt, a belly band, a pregnant support band, or a waist band. In some embodiments, the fetal health monitoring device may have at least three electrodes arranged adjacent to each other such that one measurement electrode is located on the right side of the abdomen (correspondingly, the left side), one reference electrode is located on the left side of the abdomen (correspondingly, the right side) and one offset electrode is in the middle. In some embodiments, the device of fig. 7 may have a fourth electrode positioned 90 degrees from the linear arrangement in the center of the abdomen. This fourth electrode may provide a measurement of the biopotential signal, the motion signal, the fHR signal and/or the fHRV signal in the vertical direction. In some embodiments of the present invention, the,

the device of fig. 7 may have a fifth electrode located on the back of the pregnant woman and providing a signal without uterine activity and/or maternal movement but carrying physiological and recorded artifacts that can be used to process the biopotential signal to obtain cleaner and more accurate EHG, mcg, fECG, fHR, fHRV and/or fetal movement signals.

Fig. 8 shows another embodiment of a fetal health monitoring device in which the electrode patch 46 and sensor module 18 may be integrated in a daily living accessory, which may be integrated in a pillow or cover.

In other embodiments, such as the embodiment of fig. 12, the electrode patch 50 includes a fetal health sensor module 18 that includes a plurality of sensors 210 (e.g., accelerometers, gyroscopes, and/or magnetometers), such as at least 5 sensors. In some embodiments, as shown in fig. 12, the electrode patch 50 has three ends or lobes 212, each containing a sensor 210. The additional sensor 210 is disposed at the entire middle portion of the electrode patch 50. As will be apparent from the description elsewhere herein, sensor location and number are critical to accurately identifying fetal movement, fHR and/or fHRV. In some cases, an embodiment may contain less than five sensors or more than five sensors, e.g., another sensor as a reference sensor. In this case, when a system including the fetal health sensor module 18 incorporating the plurality of sensors 210 is used, the overall motion can be identified with high efficiency.

As can be appreciated from fig. 4-8 and 12, the fetal health monitoring device is integrated in a small and easy to use form factor that does not require the operation of a clinical staff. In other words, the fetal health monitoring device is advantageously implemented in such a way that the pregnant woman may operate the device himself. Small size and extreme miniaturization can be achieved by low power electronic system design, i.e., a combination of low power circuit design, low power architectural design, and firmware optimization. Low power system design allows the battery size to be minimized and thus the overall system can be made to a very small size. Ease of use may come from a combination of intelligent electronic devices and high levels of integration. Using intelligent electronics, the device may be automatically opened when positioned on the body, or the device may automatically detect contractions, fetal movements, fHR, mHR and/or fHRV and trigger feedback accordingly, or the system may automatically detect a specific situation-e.g. the fact that the pregnant woman is moving-and adapt its signal processing accordingly. At high integration levels, the electrode patch can integrate all cables onto the electrodes and provide a very simple way for a user to connect the sensor to the electrode patch. Connecting the electrode patch to the sensor module may be done by a magnetic interface, by a snap mechanism, by a slide mechanism, by a screw mechanism or any other mechanism providing good mechanical and electrical contact between the sensor module and the electrode patch.

The use of electrode patches improves the reliability of fetal health monitoring since the user is unlikely to misplace different electrodes relative to each other, since the electrodes are always in the same relative position. The use of an electrode patch improves the experience and ease of use of fetal health monitoring as it does not require the attachment of multiple electrodes to the abdomen, but only a single electrode patch.

The device may be designed so that the pregnant woman knows how to wear the device and where to place it. The device may be designed such that it is easy to wear. For example, a pregnant woman need only pick up the sensor module, attach the sensor module to the electrode patch, and wear the sensor module.

In some embodiments, the electrode patch includes at least two electrodes. In an alternative embodiment of the apparatus, the electrode patch may contain a third electrode which may be used to bias the signal acquisition electronics to the body voltage or to apply a common mode voltage to the body to reduce measurement noise, this measurement principle also being referred to as right leg drive. In another alternative embodiment of the device, the electrode patch may comprise further measuring electrodes. Multiple measurement electrodes may be positioned at different locations of the abdomen, advantageously providing multi-dimensional measurements of uterine electrical activity, fetal movement, fHR, mHR and/or fHRV. The electrodes may or may not contain a conductive gel. Conductive gels can be used to improve the quality of contact between the body and the electrodes. The electrode patch may or may not be adhesive.

Method of producing a composite material

Some or all of the above components, or additional or alternative components, may function to monitor or determine fetal health. Some of the methods employed for monitoring or detecting fetal health are described below.

As shown in fig. 9, a computer-implemented method 100 for longitudinally monitoring fetal health during pregnancy for one embodiment comprises: acquiring a signal from a sensor S110; processing the signal to identify and extract a parameter of interest from the signal S120; and analyzing the parameter of interest to determine the fetal health S130. The method is used to monitor and/or determine fetal health over time. In some embodiments, the method is used to determine fetal health by comparing a parameter of interest to a fetal health index (fig. 11) or a personalized or population level fetal health trend (fig. 9). The method is used in the field of maternal and/or fetal health, but may additionally or alternatively be used in any suitable application, clinical or otherwise.

In some embodiments, the time span of fetal health may be within a time interval of a few seconds, such as an interval of less than 8 seconds, 4 seconds, 2 seconds, 1 second, or 0.5 seconds. In some cases, fetal movement may be averaged over a longer time interval and captured over a shorter time interval. In some cases, a longer time interval (e.g., 4 seconds) may be set, and is long enough to average the accelerations due to fetal kicks, but short enough to limit processing delays and limit maternal movement that may affect the algorithm output over the longer time interval. In some cases, the variable length features may reduce false positives. In some embodiments, low complexity time domain features such as mean, standard deviation, quartering distance, correlation between axes, sum, minimum, maximum, and magnitude may be implemented on the embedded device. In some cases, each feature may be calculated in terms of axis, sensor, and/or window time interval size.

In some embodiments, feature classification may be performed using a random forest. In some cases, the low-complexity temporal features may not be selected prior to classification, in which case a random forest may select a subset of the available low-complexity temporal features at each iteration. In some cases, the number of features may be set to be chosen at each iteration as the square root of the total number of features to maintain all information in the training phase relative to other feature selection methods. In such cases where the kick count is small relative to the total available data (i.e., the total kick for the fetus and parturient), the class imbalance may be addressed by allowing the random forest classifier to select a subset of samples during training. The best ratio between the reference category (i.e., kicking) and the majority category (i.e., non-kicking) may be determined by cross-validating and optimizing the F-score (e.g., selecting the ratio that displays the best F-score). In some cases, all data in the minority class and one fifth of the data in the majority class may be included to provide the best balance.

As shown in FIG. 9, one embodiment of a computer-implemented method 100 for longitudinally monitoring fetal health during pregnancy includes block S110, which recites acquiring signals from sensors. Block S110 is for measuring characteristics or characteristics of the fetus (e.g., fetal heart rate, maternal heart rate, heart rate variability, exercise, kick count, etc.) or the environment surrounding the fetus (e.g., amniotic fluid volume, amniotic fluid pH, placental oxygenation, etc.). For example, acquiring the signal may include: acquiring one or more signals indicative of: fetal movement, cardiac electrical activity, heart sounds, heart rate variability, amniotic fluid volume, placental oxygenation, placental temperature, placental pH, respiration, location, distress, and/or any other characteristic or feature of interest. In various embodiments, the one or more signals are sensed by a sensor having a plurality of electrodes and recorded by a processor into a memory.

As shown in FIG. 9, one embodiment of a computer-implemented method 100 for longitudinally monitoring fetal health during pregnancy includes block S120 which recites processing the signals to identify and extract a parameter of interest from the signals. Block S120 provides for isolating one or more parameters of interest from signals generated by the sensors. For example, the method may comprise: the sensor signal is amplified, filtered, digitized, and/or otherwise processed to isolate the readable signal from the acquired noise signal. The method may comprise identifying and/or extracting a parameter or series of parameters of interest from the sensor signal. The parameter of interest may be, for example, one or more of the following: average fHR, average fHRV, average fetal heart beat, fetal kick count, fetal movement count, average placental oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, fHR curve, fHRV curve, and fetal movement curve. In some embodiments, the method includes calculating a mean, median, percentile, standard deviation, or other meaningful statistic of the parameter of interest. The parameter of interest may be a physiological parameter (e.g. heart rate, amniotic fluid volume, placental oxygenation, etc.) and/or a behavioral parameter (e.g. kick count, fetal movement curve, etc.). In some embodiments, the mother-to-be may supplement the data by entering observed parameters of interest (e.g., observed kick counts, fetal position or orientation, other general feelings about the fetus, or information obtained by the pregnant woman from a doctor's follow-up) into a user interface of the mobile computing device.

In some embodiments, as in block S120, the feature or parameter of interest may be fHR. fHR may be detected by a signal from a sensor. In some cases, after pre-processing the detected signal, maternal cardiac peaks can be detected and removed from the signal. In this case, fetal heart peaks can be detected from the filtered signal to remove low frequency fluctuations as well as high frequency artifacts and noise. In some embodiments, the data (i.e., heart rate data) may be convolved with a wavelet function to emphasize peaks in the frequency bands, and then the absolute values of the resulting signals may be generated from the data. In some cases, a shorter sliding time interval may be used to analyze the signal. In some cases, the peak may be determined by reaching a threshold that conveys a signal-to-noise ratio. Once the peak is determined, a new analysis interval may be defined. In other cases, the current analysis time interval may be increased. The peak calculated can be determined using wavelet power and then optimized in the time domain.

In some embodiments, the determined maternal cardiac peak may be removed from the signal to determine a fetal cardiac peak, which may be of a smaller magnitude. In some cases, in order to remove the maternal cardiac peak, a template using the last determined maternal cardiac peak may be generated. A template may be fitted to the current peak and the adjusted template from the data may be removed. In some cases, a sub-component analysis procedure may be implemented in order to improve the template. In this case, the principal component analysis can find the principal component that represents most of the information in the maternal section. The first component may be an average of the signal and the other components may represent variations of the data relative to the first component. In this case, the first principal component may be fitted to the data.

In some embodiments, once the maternal cardiac peak is removed, the cardiac peak detection algorithm as described above can be completed again by adding wavelets centered at higher frequencies, generating the fetal cardiac peak.

As shown in FIG. 9, one embodiment of a computer-implemented method 100 for longitudinally monitoring fetal health during pregnancy includes block S130, which recites analyzing a parameter of interest to determine a fetal health level. Block S130 provides for individually evaluating or aggregating multiple parameters of interest to determine fetal health. In some embodiments, thresholding, regression models, and/or machine learning algorithms can be used to determine the probability of fetal health or distress, as described in further detail elsewhere herein.

In some embodiments, analyzing the parameter of interest may include detecting a fetal kick count by reducing false positives to improve accuracy. In some embodiments, the data set acquired by the signals from the sensors may be divided into a training set and a validation set. In some cases, the training set may contain at least two-thirds of the acquired data. In some cases, the acquired data may be randomly sampled and the validation set may contain at least one third of the acquired data. For example, 60 records may be used for the training set and 28 records may be used for the validation set. In some embodiments, the data classification may be organized as a binary classification problem that determines fetal kicking from non-fetal kicking (i.e., non-motion, noise, etc.). In some cases, due to binary classification problems and data imbalance, sensitivity and PPV may be chosen as two metrics to detect sporadic fetal kicks. In some cases, performance metrics for the entire data stream may be determined and calculated for all participants during cross-validation based on True Positives (TP), False Negatives (FN), and False Positives (FP).

In some embodiments, the performance measurement may be determined by the following equation:

sensitivity (Se): actual event record identified by the model:

positive Predictive Value (PPV): identification as an event is actually a record of the event:

where TP is true positive, FN is false negative, FP is false positive, and TN is true negative.

In some embodiments, as shown in fig. 9, the method 100 optionally includes a block S140 that recites tracking the parameter of interest over time to develop a personalized fetal health trend. The personalized fetal health trend may include measuring one or more parameters of interest hourly, daily, weekly, monthly, during every three months of pregnancy, or more or less frequently to determine a "normal" condition of the fetus on an individual basis. In one non-limiting example, the expectant mother's fetus may be healthy, but the personalized fetal health trend may show an average low fetal kick count, e.g., due to fetal size, uterine volume, amniotic fluid volume, etc. Thus, with personalized fetal health trends, the expectant mother or physician may be able to more accurately determine fetal health on an individual basis.

In some embodiments, during tracking of a parameter of interest over time, the algorithm may exhibit an increase in PPV (i.e., a reduction in false positives) when the reference accelerometer is present and when both short and long time intervals are used for feature calculation. In some cases, the algorithm may be performed at different levels, for example, calculating individual kicks at the recording level and a total kicks across the entire data set to identify the ability of the system to effectively identify individual movements within a 20 minute time interval. In some cases, the system may classify individual motions into three categories (e.g., low, medium, and high motions) and analyze the results based on the system's ability to efficiently identify the overall level of motion rather than individual motions.

In some embodiments, as shown in fig. 9, the method 100 optionally includes blocks S150 and S160 that recite: identifying deviations from personalized fetal health trends; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal distress and/or fetal health, respectively. For example, the deviation may be a parameter of interest that is above or below or otherwise deviates from the observed or measured trend of personalized fetal health. Using the above non-limiting example, if the fetal kick count falls below the average value displayed for the personalized fetal health trend, the method may comprise: it is recommended that the pregnant woman contact a healthcare provider and/or provide a probability of fetal distress.

As shown in fig. 10, a computer-implemented method 200 for longitudinally monitoring fetal health during pregnancy for one embodiment comprises: acquiring a signal from a sensor S210; processing the signal to identify and extract a parameter of interest from the signal S220; comparing the parameter of interest to a threshold S230; and determining the fetal health level S230. The method is for determining fetal health using thresholding.

As shown in fig. 10, a computer-implemented method 200 for longitudinally monitoring fetal health during pregnancy optionally includes blocks S230 and S240, which respectively recite: comparing the parameter of interest to a threshold; and determining the fetal health level. The threshold may be based on historical medical data, community data, personal data, or other empirical data. In some variations, the probability of fetal health is higher if the parameter of interest is above a threshold. Alternatively, if the parameter of interest is below a threshold, the probability of fetal distress is higher. For example, for a certain amount of amniotic fluid, the threshold may be set to the 50 th percentile, depending on the number of weeks of pregnancy (e.g., according to the profile of the expectant mother in the system). In some embodiments, if the probability indicates that the fetus is likely to be embarrassed, the method may comprise: notifying the medical care provider of the expectant mother; recommending a course of action to the expectant mother (e.g., relax, contact with a healthcare provider, drink more water, etc.); providing feedback to the mother-as-intended (e.g., relaxation, walking, etc.); and/or automatically connecting the mother-to-be with a healthcare provider or specialist. Further, the method may comprise: a degree of certainty is determined that is consistent with the determined probability.

In some embodiments, the method comprises: one or more regression models are used to analyze the parameters of interest and/or determine the probability of fetal health or distress. The regression model is used to predict one variable from one or more other variables. For example, the one or more other variables may be derived from and/or form part of a personalized fetal health trend, a population-level fetal health trend, or any other previously or simultaneously acquired or measured signal.

In some embodiments, the method comprises: machine learning is used to analyze parameters of interest and/or to determine the probability of fetal health or distress. Machine learning uses algorithms (e.g., generalized linear models, random forests, support vector machines, etc.) to make predictions, e.g., predictions about fetal health, based on one or more measured and/or analyzed signals.

As shown in fig. 11, a computer-implemented method 300 for longitudinally monitoring fetal health during pregnancy for one embodiment comprises: acquiring a signal from a sensor S310; processing the signal to identify and extract a parameter of interest from the signal S320; optionally, tracking a plurality of parameters of interest over time at a population level S330; optionally, forming a fetal health index based on the tracked parameter of interest S340; optionally, the extracted parameter of interest is compared to the fetal health index to determine the fetal health level S350. The method is used to determine fetal health by comparing one or more parameters of interest to population level data, community data, historical data, predictive data, empirical data, or any other available data. For example, as shown at block S330, one or more parameters of interest are tracked over time in a community, group of mothers, and/or any other environment to develop, generate, or otherwise establish a fetal health index.

As shown in fig. 11, the computer-implemented method 300 for longitudinally monitoring fetal health during pregnancy of an embodiment optionally includes block S340, which recites forming a fetal health index based on the tracked parameter of interest. For example, the fetal health index may comprise a numerical range, e.g., 1-10, of one or more parameters, wherein each integer, fraction, or decimal in the numerical range is associated with a percentile, a historical observation of the degree of fetal health (e.g., kick count related to fetal health, fetal movement related to fetal health, etc.), a measured degree of a fetal feature or characteristic (e.g., heart rate variability related to fetal health, fetal respiration, etc.), an amount of a measured parameter (e.g., an amount of amniotic fluid related to fetal health, placental oxygenation related to fetal health, etc.), or any other parameter.

As shown in fig. 11, the computer-implemented method 300 for longitudinally monitoring fetal health during pregnancy of an embodiment optionally includes block S350 that recites comparing the extracted parameter of interest to a fetal health index to determine a fetal health level. Block S350 provides for assessing fetal health by comparing the one or more parameters of interest to the fetal health index.

In one non-limiting example, the sensor measures the number of fetal kicks over two hours and compares the measured kick count to the fetal health index. In one embodiment of the fetal health index, each integer corresponds to a kick in a kick count assessment performed using the systems and methods described herein. The presence of zero to four kicks within two hours may indicate fetal distress, the presence of five to nine kicks within two hours may indicate a high probability of fetal health, and the presence of ten or more kicks within two hours may indicate fetal health. In some embodiments, the method comprises: if the kick count is between five and nine times over two hours, the mother-to-be is recommended to take a specific action (e.g., drink cold water, lie briefly on the back or side, eat a sweet, listen to music, press one side of the abdomen, etc.) to wake up the baby. In some embodiments, the method comprises: if, for example, the kick count is below nine times over two hours, the mother-to-be is advised to contact the healthcare provider.

In another non-limiting example, each integer corresponds to a degree of fHRV measured using the systems and methods described herein. For example, a numerical range of 0 to 3 may be used in the fetal health index, where: 0 indicates no amplitude detected and no fHRV is present; 1 denotes an amplitude of five beats per minute (bpm), and fHRV is minimal; 2 indicates an amplitude between 6bpm and 25bpm and a moderate fHRV; 3 indicates that the amplitude is greater than 25bpm and fHRV is significant.

In some embodiments, the fetal health index may comprise, for example, a binary scale or indication of fetal health. In one non-limiting example, a fetus in the correct delivery orientation receives a 1, while a fetus not in the correct delivery orientation receives a 0 (e.g., hip position).

In some embodiments, two or more parameters of interest are combined to produce a fetal health index. For example, fetal movement, amniotic fluid volume, and average fHRV (e.g., during the week of pregnancy) may be combined to generate a fetal health index.

In some embodiments, the method comprises: tracking a plurality of parameters of interest over time at a population level; and forming a fetal health index based on the tracked parameter of interest. Data from multiple users of the system may be acquired over time and/or historical data sets may be analyzed to develop population level fetal health trends. Population level fetal health trends may be derived from community data in a database, such as recorded trends, rules, correlations, and/or observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users. In some such embodiments, the method comprises: tracking the parameter of interest over time, identifying a deviation from a population-level fetal health trend; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal distress and/or fetal health.

In some variations, the bias is analyzed by a machine learning algorithm. Machine learning algorithms recognize patterns, employ computational learning (e.g., learning without explicit programming) and make predictions from data (e.g., personalized data, community data, and/or group level data). Non-limiting examples of machine learning algorithms include generalized linear models, support vector machines, and random forests.

As used in the specification and in the claims, the singular form of "a" and "the" include both the singular and the plural referents unless the context clearly dictates otherwise. For example, the term "sensor" may encompass and is intended to encompass a plurality of sensors. Sometimes, the claims and disclosure may contain terms such as "plurality," "one or more," or "at least one," however, the absence of such terms is not intended to mean, and should not be construed to mean, that a plurality is not contemplated.

When the term "about" is used (e.g., to define a length or pressure) before a numerical designation or range, the term indicates an approximation that may vary by 5%, ± 1%, or ± 0.1%. All numerical ranges provided herein include the beginning and ending numbers recited. The term "substantially" indicates the majority (i.e., greater than 50%) or substantially all of the portion of a device, substance, or composition.

In block diagrams, the components shown are depicted as discrete functional blocks, but the embodiments are not limited to systems in which the functions described herein are organized as shown. The functionality provided by each of the components may be provided by software or hardware modules that are organized differently than as presently depicted, e.g., such software or hardware may be mixed, joined, duplicated, split, distributed (e.g., within a data center or geographically), or otherwise organized differently. The functions described herein may be provided by one or more processors of one or more computers executing code stored on a tangible, non-transitory machine-readable medium. In some cases, although the singular term "medium" is used, the instructions may be distributed across different storage devices associated with different computing devices, e.g., each computing device having a different subset of instructions, which is an implementation consistent with the use of the singular term "medium" herein. In some cases, a third-party content delivery network may host some or all of the information communicated over the network, in which case to the extent that the information (e.g., content) is stated to be provisioned or otherwise provided, the information may be provided by sending instructions for retrieving the information from the content delivery network.

The reader should understand that this application describes several independently useful techniques. Applicants do not divide those technologies into separate patent applications, but group those technologies into a single document because their related subject matter helps to achieve economies in the application process. However, the unique advantages and aspects of this technology should not be combined. In some cases, embodiments address all of the deficiencies noted herein, but it is to be understood that the technology is independently useful, and that some embodiments address only a subset of such problems, or provide other benefits not mentioned, that will be apparent to those of skill in the art upon reading this disclosure. Certain techniques disclosed herein may not currently be claimed due to cost limitations, and may be claimed in a filing document as a continuation-in-app or by amending the claims of the present invention. Also, the abstract and the summary of the disclosure herein should not be considered to contain a comprehensive listing of all such technologies or aspects of such technologies, for any reason or space.

It should be understood, that the description and drawings are not intended to limit the inventive technique to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the inventive technique as defined by the appended claims. Further modifications and alternative embodiments of various aspects of the described technology will be apparent to those skilled in the art in view of this description. Accordingly, the specification and drawings are to be construed as illustrative only and are for the purpose of teaching those skilled in the art the general manner of carrying out the techniques of the present invention. It is to be understood that the forms of the present technology shown and described herein are to be taken as examples of embodiments. Elements and materials may be substituted for those shown and described herein, parts and processes may be reversed or omitted, and certain features of the present technology may be utilized independently, all as would be apparent to one skilled in the art after having the benefit of this description of the present technology. Changes may be made in the elements described herein without departing from the spirit and scope of the present technology as described in the following claims. The headings used herein are for organizational purposes only and are not meant to be used to limit the scope of the description.

As used throughout this application, the word "may" is used in a permissive sense (i.e., meaning "having the potential to"), rather than the mandatory sense (i.e., meaning "must"). The words "include/including" and the like mean including but not limited to. As used throughout this application, the singular forms "a", "an" and "the" include plural referents unless the content clearly dictates otherwise. Thus, for example, reference to "an element" includes a combination of two or more elements, although other terms and phrases such as "one or more" may be used with respect to one or more elements. Unless otherwise indicated, the term "or" is non-exclusive, i.e., encompasses both "and" or ". Terms describing conditional relationships, such as "responsive to X, and Y", "at X, Y", "if X, Y", "when X, Y", and the like, encompass contingencies in which a antecedent is necessary, a contingency in which a antecedent is sufficient, or contingencies in which a antecedent is a contributing result of a consequent, e.g., "state X occurring when condition Y is obtained" is generic to "X occurring only at Y" and "X occurring at Y and Z". Such conditional relationships are not limited to results that occur immediately after the antecedent was obtained, as some results may be delayed, and in the conditional statement, the antecedent is related to its successor, e.g., the likelihood of the antecedent occurring with the successor. The statement that multiple attributes or functions map to multiple objects (e.g., one or more processors perform steps A, B, C and D) encompasses both the mapping of all such attributes or functions to a subset of the attributes or functions to which a subset of all such objects and attributes or functions map (e.g., all processors perform steps a-D, respectively, and processor 1 performs step a, processor 2 performs a portion of steps B and C, and processor 3 performs a portion of step C and step D), unless otherwise noted. Further, unless otherwise specified, a statement that a value or action is "based on" another condition or value encompasses both the case where the condition or value is the only factor and the case where the condition or value is one of a plurality of factors. Unless otherwise stated, the statement that "each" instance of a set has a certain attribute should not be read to exclude the case that some otherwise identical or similar member of the larger set does not have the attribute, i.e., each does not necessarily mean each. Unless expressly indicated by the use of explicit language such as "perform Y after performing X," for example, limitations relating to the order of enumerated steps should not be construed as imposing a non-existent meaning to the claims, as opposed to statements such as "perform X on an item, perform Y on an item past X," which are used to render the claims more readable, rather than specifying an order, which might not be inappropriately referred to imply order limitations. A statement that references "A, B and at least Z of C," etc. (e.g., "A, B or at least Z of C") refers to at least Z of the listed categories A, B and C without requiring at least Z units in each category. Unless specifically stated otherwise, as apparent from the discussion, it is appreciated that throughout the description, discussions utilizing terms such as "processing," "computing," "calculating," "determining," or the like, refer to the action and processes of a specific apparatus, such as a special purpose computer or a similar special purpose electronic processing/computing device.

As used herein, the term "comprising" is intended to mean that the systems and methods include the recited elements, and may additionally include any other elements. "consisting essentially of … …" shall mean that the system and method includes the recited elements and excludes other elements having significance to combinations performed to achieve the stated objectives. Thus, a system or method consisting essentially of elements as defined herein would not exclude other materials, features or steps having no material effect on one or more of the basic and novel characteristics of the claimed invention. "consisting of … …" shall mean that the system and method includes the recited elements and excludes elements or steps that are not merely trivial or insignificant. Embodiments defined by each of these transitional terms are within the scope of this disclosure.

The examples and illustrations contained herein show by way of illustration, and not by way of limitation, specific embodiments in which the subject matter may be practiced. Other embodiments may be utilized and derived therefrom, such that structural and logical substitutions and changes may be made without departing from the scope of this disclosure. Such embodiments of the inventive subject matter may be referred to herein, individually or collectively, by the term "invention" merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept if more than one is in fact disclosed. Thus, although specific embodiments have been illustrated and described herein, any arrangement calculated to achieve the same purpose may be substituted for the specific embodiments shown. This disclosure is intended to cover any and all adaptations or variations of various embodiments. Combinations of the above embodiments, and other embodiments not specifically described herein, will be apparent to those of skill in the art upon reviewing the above description.

The technique of the invention will be better understood with reference to the examples listed below:

1. a system for monitoring fetal health over time during pregnancy, the system comprising: a sensor coupled to the pregnant woman; a processor communicatively coupled with the sensor; and a computer readable medium having stored thereon non-transitory processor-executable instructions, wherein execution of the instructions causes the processor to perform a method comprising: acquiring a signal from the sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health.

2. The system of embodiment 1, wherein the method performed by the processor further comprises comparing the parameter of interest to a fetal health index.

3. The system of any of embodiments 1-2, wherein the method performed by the processor further comprises tracking the parameter of interest over time to form a personalized fetal health trend.

4. The system of any of embodiments 1-3, wherein the method performed by the processor further comprises: identifying deviations from the personalized fetal health trend; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

5. The system of any of embodiments 1-4, wherein the method performed by the processor further comprises: tracking the parameter of interest over time; identifying deviations from population-level fetal health trends; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

6. The system of any of embodiments 1-5, wherein analyzing the deviation is performed by one of thresholding, a machine learning algorithm, and regression modeling.

7. The system of any of embodiments 1-6, wherein the machine learning algorithm comprises one or more of a generalized linear model, a support vector machine, and a random forest.

8. The system of any of embodiments 1-7, wherein the population-level fetal health trend is derived from community data in a database.

9. The system of any of embodiments 1-8, wherein the community data comprises recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

10. The system of any of embodiments 1-9, wherein the system comprises a plurality of sensors.

11. The system of any of embodiments 1-10, wherein acquiring a signal comprises acquiring a plurality of signals.

12. The system of claim 1, wherein a plurality of parameters are extracted.

13. The system of any of embodiments 1-11, wherein the plurality of parameters comprises physiological, activity, and behavioral parameters.

14. The system of any of embodiments 1-12, wherein the sensors comprise one or more sensors configured to measure one or more of: fetal movement, fetal cardiac electrical activity, fetal heart sounds, fHR, fHRV, amniotic fluid volume, placental oxygenation, placental temperature, placental pH, fetal respiration, fetal position, fetal orientation, and fetal distress.

15. The system of any of embodiments 1-14, wherein the sensor senses one or more of a biopotential signal, an inertial signal, an acoustic signal, a bioimpedance signal, an optical signal, a near infrared spectroscopy signal, an electrochemical signal, and a temperature signal.

16. The system of any of embodiments 1-15, wherein the parameter of interest comprises one or more of: average fHR, average fetal heart beat, fetal kick count, fetal movement count, average placental oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, fHR curve, and fetal movement curve.

17. The system of any of embodiments 1-16, further comprising a portable wearable sensor patch comprising the sensor, the processor, and the computer-readable medium.

18. The system of any of embodiments 1-17, wherein the wearable sensor patch further comprises a wireless antenna for communicating with a mobile computing device.

19. The system of any of embodiments 1-18, wherein the sensor is located on or in a portable wearable sensor patch, the sensor patch further comprising electronic circuitry and a wireless antenna, and wherein the sensor patch is in wireless communication with a mobile computing device comprising the processor and the computer-readable medium.

20. The system of any of embodiments 1-19, wherein the method performed by the processor further comprises one or more of: generating an alert, providing feedback to the pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

21. The system of any of embodiments 1-20, wherein the method performed by the processor further comprises notifying a healthcare provider of the fetal health level.

22. The system according to any of embodiments 1-21, wherein the method performed by the processor further comprises determining a probability of fetal distress.

23. The system of any of embodiments 1-22, wherein the method performed by the processor further comprises determining a degree of certainty that is consistent with the determined probability.

24. The system of any of embodiments 1-23, wherein the method performed by the processor further comprises determining a probability that a fetus is healthy.

25. The system of any of embodiments 1-24, wherein the method performed by the processor further comprises determining a degree of certainty that is consistent with the determined probability.

26. The system of any of embodiments 1-25, wherein the step of analyzing the parameter of interest further comprises: comparing the parameter of interest to a threshold.

27. The system of any of embodiments 1-26, wherein the probability of fetal health is higher if the parameter of interest is above the threshold.

28. The system according to any of embodiments 1-27, wherein the probability of fetal distress is higher if the parameter of interest is below the threshold.

29. The system of embodiment 28, wherein the step of analyzing the parameter of interest further comprises: analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

30. A computer-implemented method for longitudinally monitoring fetal health during pregnancy outside of a hospital environment, the method comprising: acquiring a signal from a sensor; processing the signal to identify and extract a parameter of interest from the signal; and analyzing the parameter of interest to determine fetal health.

31. The method of embodiment 30, further comprising comparing the extracted parameter of interest to a fetal health index.

32. The method of any of embodiments 30-31, further comprising tracking the parameter of interest over time to develop personalized fetal health trends.

33. The method of any of embodiments 30-32, further comprising: identifying deviations from the personalized fetal health trend; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

34. The method of any one of embodiments 30-33, further comprising: tracking the parameter of interest over time; identifying deviations from population-level fetal health trends; and analyzing the deviation to determine whether the deviation is indicative of a change in fetal health and/or fetal distress.

35. The method of any of embodiments 30-34, wherein analyzing the deviation is performed by a machine learning algorithm.

36. The method of any of embodiments 30-35, wherein the machine learning algorithm comprises one or more of a generalized linear model, a support vector machine, and a random forest.

37. The method of any of embodiments 30-36, wherein the population-level fetal health trend is derived from community data in a database.

38. The method of any of embodiments 30-37, wherein the community data comprises recorded trends, rules, associations, and observations generated by tracking, aggregating, and analyzing one or more physiological, biological, or activity parameters from a plurality of users.

39. The method of any one of embodiments 30-38, further comprising acquiring a plurality of signals.

40. The method of any one of embodiments 30-39, further comprising extracting a plurality of parameters of interest.

41. The method of any of embodiments 30-40, wherein the sensor senses one or more of a biopotential signal, an inertial signal, an acoustic signal, a bioimpedance signal, an optical signal, a near infrared spectroscopy signal, an electrochemical signal, and a temperature signal.

42. The method of any of embodiments 30-41, wherein the parameter of interest comprises one or more of: average fHR, average fetal heart beat, fetal kick count, fetal movement count, average placental oxygenation level, average placental temperature, average placental pH, average amniotic fluid volume, fHR curve, and fetal movement curve.

43. The method of any one of embodiments 30-42, further comprising one or more of: generating an alert, providing feedback to a pregnant woman, recommending an action to the pregnant woman, and automatically connecting the pregnant woman with a healthcare provider.

44. The method of any one of embodiments 30-43, further comprising notifying a healthcare provider of the fetal health level.

45. The method according to any one of embodiments 30-44, further comprising determining a probability of fetal distress.

46. The method of any of embodiments 30-45, further comprising determining a degree of certainty that is consistent with the determined probability.

47. The method of any one of embodiments 30-46, further comprising determining a probability that a fetus is healthy.

48. The method of any one of embodiments 30-47, further comprising: a degree of certainty is determined that is consistent with the determined probability.

49. The method of any of embodiments 30-48, wherein the step of analyzing the parameter of interest further comprises: comparing the parameter of interest to a threshold.

50. The method according to any of embodiments 30-49, wherein the probability of fetal health is higher if the parameter of interest is above the threshold.

51. The method according to any of embodiments 1-50, wherein the probability of fetal distress is higher if the parameter of interest is below the threshold.

52. The method according to any one of embodiments 30-51, further comprising analyzing the parameter of interest using a regression model or a machine learning algorithm to determine a probability of fetal health or distress.

53. The method of any one of embodiments 30-52, further comprising: tracking a plurality of the parameters of interest over time at a population level; and forming a fetal health index based on the tracked parameter of interest.

54. The method of embodiment 53, further comprising: comparing the extracted parameter of interest to the fetal health index to determine a fetal health level.