阅读说明：本技术 一种排尿预测方法及装置 (Urination prediction method and device ) 是由 彭伟鸿 李笑 曾梓轩 于 2019-10-29 设计创作，主要内容包括：本发明公开了一种排尿预测方法及装置,针对患者无法感知尿意或无意识排尿的情况,采取测量及预测机制,通过测量患者储尿阶段的阻抗值预测得到患者未来的排尿时间,从而实现提醒患者的膀胱需适时排尿的有效性。本发明可以针对不同年龄段、体重的患者进行准确预测,同时基于生物阻抗分析技术进行膀胱阻抗值的采集,在不对患者造成创口的情况下能够预估患者的膀胱储尿情况,简化了排尿预测程序。(The invention discloses a urination prediction method and a device, aiming at the condition that a patient cannot sense the urination intention or unconscious urination, a measurement and prediction mechanism is adopted, and the future urination time of the patient is predicted by measuring the impedance value of the urine storage stage of the patient, so that the effectiveness of reminding the bladder of the patient of needing to urinate at the right time is realized. The invention can accurately predict patients with different ages and weights, and simultaneously, the bladder impedance value is collected based on the bioimpedance analysis technology, so that the bladder urine storage condition of the patient can be predicted under the condition of not causing wounds to the patient, and the urination prediction program is simplified.)

1. A urination prediction method, comprising the steps of:

s1, collecting data required by urination prediction, comprising the following steps: physical parameters of testers and patients related to urination, the daily urination frequency of the testers and the bladder impedance value of the patients in each time period;

s2, classifying the acquired data to respectively obtain a data set for predicting the urination times of a patient in one day and a data set for predicting the urination time of the patient;

s3, arranging the urination times of the testers according to the adjacent relation through the adjacent relation between the physical parameters of the testers and the patients on the basis of a data set for predicting the urination times of the patients in one day, and solving the average value of the urination times of the patients, namely the predicted urination times of the patients;

s4, based on a data set for predicting the urination time of the patient, randomly selecting a certain moment and the bladder impedance value corresponding to the certain moment as a mark point, using the other moments and the bladder impedance values corresponding to the other moments as data points, calculating to obtain the adjacent relation between a plurality of data points and the mark point, arranging the bladder impedance values of the patient and the corresponding moments according to the adjacent relation, and respectively calculating the bladder impedance value of the patient and the average value of the corresponding moments as prediction points; repeatedly selecting mark points at different moments to obtain a prediction point to obtain an impedance prediction curve, and performing inverse function transformation on the impedance prediction curve to obtain the predicted urination time of the patient;

and S5, forming a prediction set by the predicted urination times and the predicted urination time of the patient obtained in the steps S3 and S4, predicting the urination times and the urination time of the patient by using the prediction set, and outputting a prediction result.

2. The urination prediction method as claimed in claim 1, wherein the physical parameters related to urination in step S1 include the ages and BMI indexes of the testers and patients.

3. The urination prediction method as claimed in claim 1, wherein the bladder impedance values of the patient at each time interval in step S1 are measured by connecting the test electrodes to the test site of the patient by a four-electrode connection method based on a bio-impedance analysis technique, and the measured bladder impedance values are pre-processed to remove abnormal points.

4. The urination prediction method according to claim 2, wherein the step S2 specifically includes:

defining the age of the subject or patient and the BMI index data as influence values;

taking the influence value of the tester, the urination frequency of the tester, the bladder impedance value of the patient and the bladder impedance value acquisition time as a data set { ASIT };

the matrix { A S } is used as a data set for predicting the urination times of a patient in one day, and the matrix { I T } is used as a data set for predicting the urination time of the patient;

wherein the matrix of influence values is a ═ age BMI

The micturition time matrix is: s ═ times of urination S

The time period bladder impedance value matrix is: i ═ I_t1i_t2i_t3... i_tn]，i_tnDenotes the t-th_nA period of time t_nThe bladder impedance value matrix at m moments in time periods is:

i_tnmdenotes the t-th_nThe mth moment in time

The time period matrix is: t ═ T₁t₂t₃... t_n]

Wherein the t th_nThe time matrix of m times in each time segment is:

t-th in the above formula_nThe time periods are the time periods when the patient needs to urinate.

5. The urination prediction method according to claim 4, wherein the step S3 specifically includes:

the impact and number of urination for l subjects were recorded and the matrix { A S } was expanded to:

setting age of patient as age_preBMI index is BMI_pre；

Selecting the patient parameter r from the matrix A_pre＝(age_pre,BMI_pre) Adjacent k_sPoint r_j＝(age_j,BMI_j) Calculating the point r by using Euler distance formula_preTo adjacent k_sPoint r_jThe distance of (c):

wherein r is_pre＝(age_pre,BMI_pre)，r_j＝(age_j,BMI_j)；

The distance values obtained by calculation are arranged from small to large to obtain k_sA sorting sign matrix d formed by distance values_s

Based on the classification marks, the times of urination of the represented testers are arranged to obtain a urination time matrix

To urination time matrix

wherein the urination time matrix

to urination time matrix

wherein the weighted value

6. The urination prediction method according to claim 5, wherein the step S4 specifically includes:

normalizing the bladder impedance values according to the t-th time period_nBladder impedance value matrix i at m moments in time period_tnThe bladder impedance value matrix at m moments in the first time period is as follows:

according to the t_nTime matrix t of m time of each time segment_nThe time matrix of m times in the first time period is:

in the above matrix i_t1And matrix t₁In which a marking point z is arbitrarily selected according to the corresponding relation_m＝(i_t1m,t_1m) At the marked point z_mSelecting adjacent k_tData points z_j＝(i_kj,t_kj) As sample points;

calculating the marking point z by using a distance formula_mTo k_tDistance d between sample points_sj；

The obtained distance value d_sjArranged from small to large to obtain k_tThe distance value forms a classification mark d_t

Based on the class label d_tThe bladder impedance values represented by the k are ranked to obtain k_tMatrix i of individual bladder impedance values_kt1Comprises the following steps:

k_ttime matrix t corresponding to bladder impedance value_kt1Comprises the following steps:

according to matrix i_kt1Calculate k_tThe average or weighted average of the bladder impedance values is taken as the predicted bladder impedance value i'_t11；

Wherein k is obtained_tPredicted bladder impedance value i 'obtained by averaging bladder impedance values'_t11Comprises the following steps:

finding k_tPredicted bladder impedance value i 'obtained from weighted average of bladder impedance values'_t11Comprises the following steps:

wherein the weighted value

by matrix t_kt1Finding k_tThe average value of each time is used as the predicted time t'₁₁

The first data point w predicted by the above steps_t11＝(i′_t11,t′₁₁)

In the above matrix i_t1And matrix t₁In the method, a second marking point z is randomly selected according to the corresponding relation_m′＝(i_t1m′,t_1m′) Calculating a predicted second data point w_t12＝(i′_t12,t′₁₂)

And recalculating the predicted bladder impedance values at m' moments of the first time period:

the bladder impedance values at m' times correspond to the following times:

wherein m' is less than or equal to m;

for the predicted bladder impedance value i_t1preAnd corresponding time t_1preConnecting the formed points to obtain a prediction curve W of a first time period_t1It is compared with the measured bladder impedance value i_t1Making a comparison, i.e. predicting the curve W_t1Upper bladder impedance value and i_t1The bladder impedance values in the bladder are compared by adopting mean square error, and the obtained error is as follows:

if the obtained error is within the preset error range, the predicted curve W is represented_t1The fitting is accurate; otherwise, k is reselected_tCalculating the value of (1) until the error is within a preset error range, and stopping calculating;

for matrix i in the second time period_t2And t₂The steps are adopted to obtain a prediction curve W of the second time period_t2(ii) a Analogizing in turn to obtain the prediction curve of each time segmentW：

W＝[W_t1W_t2... W_tn]

And carrying out nonlinear fitting on the prediction curve W to obtain an impedance prediction curve W (i, t), and carrying out inverse function transformation on the impedance prediction curve W to obtain:

W_pre＝W(t,i)

based on W obtained by performing inverse function transformation_preTaking t_nTime t of one time segment_nmBladder impedance value i_tnmObtaining the predicted urination time of the patient:

t_pai＝θ·W_pre(t,i_tnm)

where θ is an influence coefficient, the magnitude of which is determined by the influence value.

7. The urination prediction method as set forth in claim 6, wherein the marked point z is calculated by a distance formula in step S4_mTo k_tDistance d between sample points_sjCalculated by any of the following formulas:

a,

Di, dt_j＝|z_m-z_j|

III,

fourthly,

V, V,

Z in the above formula_m＝(i_t1m,t_1m)，z_j＝(i_kj,t_kj)。

8. An apparatus for a method of predicting urination, comprising:

the data acquisition module is used for acquiring data required by urination prediction and comprises: physical parameters of testers and patients related to urination, the daily urination frequency of the testers and the bladder impedance value of the patients in each time period;

the data classification module is used for classifying the acquired data to respectively obtain a data set for predicting the urination times of a patient in one day and a data set for predicting the urination time of the patient;

the urination frequency predicting module is used for arranging the urination frequencies of the testers according to the adjacent relation between the physical parameters of the testers and the patients on the basis of a data set for predicting the daily urination frequency of the patients, and solving the average value of the urination frequencies of the patients as the predicted urination frequency of the patients;

the urination time prediction module is used for randomly selecting a certain moment and the bladder impedance value corresponding to the certain moment as a mark point and the other moments and the bladder impedance values corresponding to the other moments as data points based on a data set for predicting the urination time of the patient, calculating the adjacent relation between a plurality of data points and the mark point, arranging the bladder impedance values of the patient and the corresponding moments according to the adjacent relation, and respectively calculating the bladder impedance value of the patient and the average value of the corresponding moments as prediction points; repeatedly selecting mark points at different moments to obtain a prediction point to obtain an impedance prediction curve, and performing inverse function transformation on the impedance prediction curve to obtain the predicted urination time of the patient;

and the prediction result output module is used for forming a prediction set by the predicted urination times and the predicted urination time of the patient obtained by the urination time prediction module and the urination time prediction module, predicting the urination times and the urination time of the patient by using the prediction set and outputting a prediction result.

Technical Field

The invention relates to the technical field of prediction of physiological events, in particular to a urination prediction method and device.

Background

The neurogenic bladder patient refers to bladder and urethra dysfunction caused by damage of functional nerves controlling urination or peripheral nerves in a central nervous system, and the patient usually loses consciousness of sensing urine storage and urination, so that complications such as urinary system infection and the like are caused, and the physical and psychological health of the patient is greatly influenced; the urinary incontinence patient cannot predict the urination time, the urination is unconscious, and the urine is always leaked involuntarily when the urination is happened, which brings serious afflictions to the physiology and psychology of the patient. Therefore, it is necessary to predict urination time of a patient with bladder dysfunction such as neurogenic bladder or urinary incontinence, to take an auxiliary urination measure or an autonomous urination measure in advance, and to avoid or reduce the influence of complications such as urinary system infection.

The existing urination prediction method generally adopts ultrasound, pressure and displacement to detect bladder capacity so as to predict urination, the sampling modes of the methods are complex to operate, and some of the methods can cause wounds to patients; and because these modes are difficult to realize the real-time monitoring to the bladder, the effectiveness of reminding the bladder of the patient to urinate at the right time cannot be realized.

Disclosure of Invention

The invention provides a urination prediction method and a device for solving the problems that the prediction process of the existing urination prediction method is inconvenient and the patient cannot be reminded of urination at proper time effectively.

In order to achieve the above purpose, the technical means adopted is as follows:

a urination prediction method comprising the steps of:

s1, collecting data required by urination prediction, comprising the following steps: physical parameters of testers and patients related to urination, the daily urination frequency of the testers and the bladder impedance value of the patients in each time period;

s2, classifying the acquired data to respectively obtain a data set for predicting the urination times of a patient in one day and a data set for predicting the urination time of the patient;

s3, arranging the urination times of the testers according to the adjacent relation through the adjacent relation between the physical parameters of the testers and the patients on the basis of a data set for predicting the urination times of the patients in one day, and solving the average value of the urination times of the patients, namely the predicted urination times of the patients;

s4, based on a data set for predicting the urination time of the patient, randomly selecting a certain moment and the bladder impedance value corresponding to the certain moment as a mark point, using the other moments and the bladder impedance values corresponding to the other moments as data points, calculating to obtain the adjacent relation between a plurality of data points and the mark point, arranging the bladder impedance values of the patient and the corresponding moments according to the adjacent relation, and respectively calculating the bladder impedance value of the patient and the average value of the corresponding moments as prediction points; repeatedly selecting mark points at different moments to obtain a prediction point to obtain an impedance prediction curve, and performing inverse function transformation on the impedance prediction curve to obtain the predicted urination time of the patient;

and S5, forming a prediction set by the predicted urination times and the predicted urination time of the patient obtained in the steps S3 and S4, predicting the urination times and the urination time of the patient by using the prediction set, and outputting a prediction result.

In the scheme, aiming at the condition that the patient cannot sense the urine or unconscious urination, a measuring and predicting mechanism is adopted, and the future urination time of the patient is predicted by measuring the physical parameters of the tester related to the urine of the patient and the impedance value of the urine storage stage of the patient.

Preferably, the physical parameters related to urination in step S1 include the ages of the testers and patients and BMI index.

Preferably, the bladder impedance value of each time period of the patient in step S1 is measured by connecting the test electrode to the test site of the patient by a four-electrode connection method based on a bioimpedance analysis technique, and the measured bladder impedance value needs to be preprocessed to remove outliers.

Preferably, the step S2 specifically includes:

defining the age of the subject or patient and the BMI index data as influence values;

taking the influence value of the tester, the urination frequency of the tester, the bladder impedance value of the patient and the bladder impedance value acquisition time as a data set { ASIT };

the matrix { A S } is used as a data set for predicting the urination times of a patient in one day, and the matrix { I T } is used as a data set for predicting the urination time of the patient;

wherein the matrix of influence values is a ═ age BMI

The micturition time matrix is: s ═ times of urination S

The time period bladder impedance value matrix is: i ═ I_t1i_t2i_t3... i_tn]，i_tnDenotes the t-th_nA period of time

Wherein the t th_nIn a time periodThe bladder impedance value matrix at m moments is:

i_tnmdenotes the t-th_nThe mth moment in time

The time period matrix is: t ═ T₁t₂t₃... t_n]

Wherein the t th_nThe time matrix of m times in each time segment is:

t-th in the above formula_nThe time periods are the time periods when the patient needs to urinate.

Preferably, the step S3 specifically includes:

the impact and number of urination for l subjects were recorded and the matrix { A S } was expanded to:

setting age of patient as age_preBMI index is BMI_pre，

Selecting the patient parameter r from the matrix A_pre＝(age_pre,BMI_pre) Adjacent k_sPoint r_j＝(age_j,BMI_j) Calculating the point r by using Euler distance formula_preTo adjacent k_sPoint r_jThe distance of (c):

wherein r is_pre＝(age_pre,BMI_pre)，r_j＝(age_j,BMI_j)；

The distance values obtained by calculation are arranged from small to large to obtain k_sA sorting sign matrix d formed by distance values_s

Based on the classification marks, the times of urination of the represented testers are arranged to obtain a urination time matrix

To urination time matrix

Calculating an average value or a weighted average value to obtain the predicted urination times of the patient;

wherein the urination time matrixThe predicted number of urination of the patient when averaged is:

to urination time matrix

The predicted number of urination of the patient when the weighted average is found is:

wherein the weighted value

d_jIs the jth distance value, and c is a preset constant value.

Preferably, the step S4 specifically includes:

normalizing the bladder impedance values according to the t-th time period_nBladder impedance value matrix i at m moments in time period_tnThe bladder impedance value matrix at m moments in the first time period is as follows:

according to the t_nTime matrix t of m time of each time segment_nThe time matrix of m times in the first time period is:

in the above matrix i_t1And matrix t₁In which a marking point z is arbitrarily selected according to the corresponding relation_m＝(i_t1m,t_1m) At the marked point z_mSelecting adjacent k_tData points z_j＝(i_kj,t_kj) As sample points;

calculating the marking point z by using a distance formula_mTo k_tDistance d between sample points_sj；

The obtained distance value d_sjArranged from small to large to obtain k_tThe distance value forms a classification mark d_t

Based on the class label d_tThe bladder impedance values represented by the k are ranked to obtain k_tMatrix i of individual bladder impedance values_kt1Comprises the following steps:

k_ttime matrix t corresponding to bladder impedance value_kt1Comprises the following steps:

according to matrix i_kt1Calculate k_tThe average or weighted average of the bladder impedance values is taken as the predicted bladder impedance value i'_t11；

Wherein k is obtained_tPredicted bladder impedance value i 'obtained by averaging bladder impedance values'_t11Comprises the following steps:

finding k_tPredicted bladder impedance value i 'obtained from weighted average of bladder impedance values'_t11Comprises the following steps:

wherein the weighted value

d_jIs the jth distance value, and c is a preset constant value;

by matrix t_kt1Finding k_tThe average value of each time is used as the predicted time t'₁₁

The predicted first data point wt11 ═ i'_t11,t′₁₁)

In the above matrix i_t1And matrix t₁In the method, a second marking point z is randomly selected according to the corresponding relation_m′＝(i_t1m′,t_1m′) Calculating a predicted second data point w_t12＝(i′_t12,t′₁₂)

And recalculating the predicted bladder impedance values at m' moments of the first time period:

the bladder impedance values at m' times correspond to the following times:

wherein m' is less than or equal to m;

for the predicted bladder impedance valueAnd corresponding time t_1preConnecting the formed points to obtain a prediction curve W of a first time period_t1It is compared with the measured bladder impedance value i_t1Making a comparison, i.e. predicting the curve W_t1Upper bladder impedance value and i_t1The bladder impedance values in the bladder are compared by adopting mean square error, and the obtained error is as follows:

if the obtained error is within the preset error range, the predicted curve W is represented_t1The fitting is accurate; otherwise, k is reselected_tCalculating the value of (1) until the error is within a preset error range, and stopping calculating;

for matrix i in the second time period_t2And t₂The steps are adopted to obtain a prediction curve W of the second time period_t2(ii) a And analogizing to obtain a prediction curve W of each time period:

W＝[W_t1W_t2... W_tn]

and carrying out nonlinear fitting on the prediction curve W to obtain an impedance prediction curve W (i, t), and carrying out inverse function transformation on the impedance prediction curve W to obtain:

W_pre＝W(t,i)

based on W obtained by performing inverse function transformation_preTaking t_nTime t of one time segment_nmBladder impedance value i_tnmObtaining the predicted urination time of the patient:

t_pai＝θ·W_pre(t,i_tnm)

where θ is an influence coefficient, the magnitude of which is determined by the influence value.

Preferably, the marked point z is calculated by using a distance formula as described in step S4_mTo k_tDistance d between sample points_sjCalculated by any of the following formulas:

a,

Di, dt_j＝|z_m-z_j|

III,

Sigma is a preset constant value;

fourthly,

V, V,

Z in the above formula_m＝(i_t1m,t_1m)，z_j＝(i_kj,tk_j)。

The present invention also provides a urination prediction device including:

the data acquisition module is used for acquiring data required by urination prediction and comprises: physical parameters of testers and patients related to urination, the daily urination frequency of the testers and the bladder impedance value of the patients in each time period;

the data classification module is used for classifying the acquired data to respectively obtain a data set for predicting the urination times of a patient in one day and a data set for predicting the urination time of the patient;

the urination frequency predicting module is used for arranging the urination frequencies of the testers according to the adjacent relation between the physical parameters of the testers and the patients on the basis of a data set for predicting the daily urination frequency of the patients, and solving the average value of the urination frequencies of the patients as the predicted urination frequency of the patients;

the urination time prediction module is used for randomly selecting a certain moment and the bladder impedance value corresponding to the certain moment as a mark point and the other moments and the bladder impedance values corresponding to the other moments as data points based on a data set for predicting the urination time of the patient, calculating the adjacent relation between a plurality of data points and the mark point, arranging the bladder impedance values of the patient and the corresponding moments according to the adjacent relation, and respectively calculating the bladder impedance value of the patient and the average value of the corresponding moments as prediction points; repeatedly selecting mark points at different moments to obtain a prediction point to obtain an impedance prediction curve, and performing inverse function transformation on the impedance prediction curve to obtain the predicted urination time of the patient;

and the prediction result output module is used for forming a prediction set by the predicted urination times and the predicted urination time of the patient obtained by the urination time prediction module and the urination time prediction module, predicting the urination times and the urination time of the patient by using the prediction set and outputting a prediction result.

Compared with the prior art, the technical scheme of the invention has the beneficial effects that:

the urination prediction method and the urination prediction device provided by the invention adopt a measurement and prediction mechanism aiming at the condition that a patient cannot sense the urination intention or unconscious urination, and predict the future urination time of the patient by measuring the impedance value of the urine storage stage of the patient, thereby realizing the effectiveness of reminding the bladder of the patient to urinate at proper time. The invention can accurately predict patients with different ages and weights, and simultaneously, the bladder impedance value is collected based on the bioimpedance analysis technology, so that the bladder urine storage condition of the patient can be predicted under the condition of not causing wounds to the patient, and the urination prediction program is simplified.

The urination detection method and the urination detection device have the advantages of safety, convenience, no side effect and low cost.

Drawings

FIG. 1 is a general flow diagram of the process of the present invention.

FIG. 2 is a flowchart illustrating steps S2-S4 of the present invention.

Fig. 3 is a BMI index-age chart for patients and individual trials.

FIG. 4 is a bladder impedance versus time plot for the period from urine storage to urination by a patient.

FIG. 5 is a block diagram of the apparatus of the present invention.

Detailed Description

The drawings are for illustrative purposes only and are not to be construed as limiting the patent;

for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product;

it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.

The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.