阅读说明：本技术 深度神经网络预测血压的训练方法、计算机装置和存储介质 (Training method for predicting blood pressure by deep neural network, computer device and storage medium ) 是由 邹丽丽 江恒炳 黄德群 于 2021-07-13 设计创作，主要内容包括：本发明公开了一种深度神经网络预测血压的训练方法,包括对脉搏波信号、心电信号和动脉血压信号进行预处理,将脉搏波信号和心电信号输入至深度神经网络模型进行多尺度融合多任务回归预测,从动脉血压信号中提取的收缩压、舒张压和平均动脉压作为期望输出,确定损失函数的值,当损失函数的值满足收敛条件时结束训练等步骤。本发明训练得到的深度神经网络模型具有根据脉搏波信号和心电信号高精度预测血压的能力,具有连续、实时、操作方便等优势,可以用于血压测量仪器的校正等非治疗用途,校正过程无需依赖专用的器具,从而向使用者提供校正的条件,方便使用者校正血压测量仪器,改善血压测量仪器的使用效果。本发明广泛应用于人工智能技术领域。(The invention discloses a training method for predicting blood pressure by a deep neural network, which comprises the steps of preprocessing a pulse wave signal, an electrocardiosignal and an arterial blood pressure signal, inputting the pulse wave signal and the electrocardiosignal into a deep neural network model for multi-scale fusion and multi-task regression prediction, taking systolic pressure, diastolic pressure and average arterial pressure extracted from the arterial blood pressure signal as expected outputs, determining the value of a loss function, finishing training when the value of the loss function meets a convergence condition and the like. The deep neural network model obtained by training has the capability of predicting the blood pressure with high precision according to the pulse wave signals and the electrocardiosignals, has the advantages of continuity, real time, convenient operation and the like, can be used for correction and other non-treatment purposes of the blood pressure measuring instrument, does not need to depend on special appliances in the correction process, thereby providing correction conditions for a user, facilitating the user to correct the blood pressure measuring instrument and improving the use effect of the blood pressure measuring instrument. The invention is widely applied to the technical field of artificial intelligence.)

1. A training method for predicting blood pressure by a deep neural network is characterized by comprising the following steps:

acquiring a pulse wave signal, an electrocardiosignal and an arterial blood pressure signal;

filtering the pulse wave signal and the electrocardio signal;

extracting systolic pressure, diastolic pressure and mean arterial pressure from the arterial blood pressure signal;

inputting the pulse wave signals and the electrocardiosignals into the deep neural network model, and extracting multi-scale features from the deep neural network model to perform multi-task regression prediction;

determining a value of a loss function from an actual output of the deep neural network model and an expected output of the deep neural network model with the systolic pressure, the diastolic pressure, and the mean arterial pressure as the expected outputs;

and when the value of the loss function meets the convergence condition, finishing the training of the deep neural network model.

2. The training method for predicting blood pressure by using the deep neural network as claimed in claim 1, wherein the acquiring of the pulse wave signal, the electrocardio signal and the arterial blood pressure signal comprises:

measuring the pulse wave signal, the electrocardiosignal and the arterial blood pressure signal;

setting time limit, amplitude peak value limit and peak time interval limit, and screening and segmenting the pulse wave signals, the electrocardiosignals and the arterial blood pressure signals.

3. The training method for predicting blood pressure by using the deep neural network as claimed in claim 1, wherein the filtering the pulse wave signal and the cardiac signal comprises:

filtering the pulse wave signal and the electrocardiosignal using a discrete wavelet transform;

setting db8 mother wavelets, wherein the decomposition layer number of the pulse wave signals is 8;

decomposing an approximate coefficient and a detail coefficient from the pulse wave signal, zeroing the approximate coefficient of the pulse wave signal at the 8 th level, zeroing the detail coefficient of the pulse wave signal at the first level, and performing soft threshold denoising and coefficient reconstruction on the pulse wave signal;

setting the number of decomposition layers of the electrocardiosignals to be 7;

decomposing an approximation coefficient and a detail coefficient from the electrocardiosignals, zeroing the 7 th level of the approximation coefficient of the electrocardiosignals, zeroing the first level of the detail coefficient of the electrocardiosignals, and performing soft threshold denoising and coefficient reconstruction on the electrocardiosignals;

thereby filtering the components of the pulse wave signals which are less than 0.25Hz and more than 31.125Hz, and filtering the components of the electrocardiosignals which are less than 0.5Hz and more than 31.125 Hz.

4. The training method for predicting blood pressure by using deep neural network as claimed in claim 1, wherein said extracting systolic pressure, diastolic pressure and mean arterial pressure from said arterial blood pressure signal comprises:

acquiring a wave peak value of the arterial blood pressure signal as the systolic pressure;

acquiring a trough value of the arterial blood pressure signal as the diastolic pressure;

obtaining a weighted average of the systolic pressure and the diastolic pressure as the mean arterial pressure; wherein the weight of the systolic pressure is 1, and the diastolic pressure is 2.

5. The training method for predicting blood pressure by using the deep neural network as claimed in claim 1, further comprising the following steps:

before the pulse wave signals and the electrocardiosignals are input into the deep neural network model, normalization processing is further carried out on the pulse wave signals and the electrocardiosignals.

6. The training method for predicting the blood pressure by the deep neural network as claimed in claim 1, wherein the deep neural network model is an MS-CNN network, the MS-CNN network comprises an input layer, a convolutional layer, a BN layer, a Relu layer, a pooling layer and a fully-connected layer, the MS-CNN network comprises a plurality of channels, and convolutional kernels of different channels are different in size so as to extract features with different scales.

7. The training method for predicting blood pressure by using the deep neural network as claimed in claim 6, wherein the inputting the pulse wave signal and the electrocardio signal into the deep neural network model, extracting multi-scale features from the deep neural network model and performing multi-task regression prediction comprises:

the deep neural network model performs convolution processing and maximum pooling processing on the pulse wave signals and the electrocardiosignals for one time;

extracting features of the pulse wave signals and the electrocardiosignals by each channel in the deep neural network model;

after the features extracted by each channel pass through the BN layer and the Relu layer, the pooling layer is used for carrying out average pooling on the features extracted by each channel in different scales;

and performing regression analysis on the features subjected to average pooling by the full connection layer to obtain the multi-task actual output of the deep neural network model.

8. The training method for predicting blood pressure of deep neural network of claim 1, wherein the loss function is MSE loss function.

9. A computer device comprising a memory for storing at least one program and a processor for loading the at least one program to perform the training method for deep neural network predictive blood pressure of any one of claims 1-8.

10. A storage medium having stored therein a program executable by a processor, wherein the program executable by the processor is configured to perform the training method for predicting blood pressure of the deep neural network according to any one of claims 1 to 8 when being executed by the processor.

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a training method for predicting blood pressure by a deep neural network, a computer device and a storage medium.

Background

The existing blood pressure measuring technology is generally based on the principles of direct measurement, arterial tension, volume compensation, oscillography, auscultation and the like, and a blood pressure measuring instrument based on the principles can directly or indirectly measure the blood pressure value. Blood pressure measuring instruments, like other gauges, also face calibration problems. In the prior art, the blood pressure measuring instrument is periodically sent to a manufacturer or a relevant organization for correction, and even professional users such as medical institutions and the like generally cannot correct the blood pressure measuring instrument due to the lack of special correction equipment.

Disclosure of Invention

In view of at least one of the above technical problems, it is an object of the present invention to provide a training method, a computer device and a storage medium for deep neural network prediction of blood pressure.

In one aspect, an embodiment of the present invention includes a training method for predicting blood pressure by a deep neural network, including:

acquiring a pulse wave signal, an electrocardiosignal and an arterial blood pressure signal;

filtering the pulse wave signal and the electrocardio signal;

extracting systolic pressure, diastolic pressure and mean arterial pressure from the arterial blood pressure signal;

inputting the pulse wave signals and the electrocardiosignals into the deep neural network model, extracting different scale features in the signals, and extracting multi-scale features from the deep neural network model to perform multi-task regression prediction on blood pressure values;

determining a value of a loss function from an actual output of the deep neural network model and an expected output of the deep neural network model with the systolic pressure, the diastolic pressure, and the mean arterial pressure as the expected outputs;

and when the value of the loss function meets the convergence condition, finishing the training of the deep neural network model.

Further, the acquiring of the pulse wave signal, the electrocardiosignal and the arterial blood pressure signal includes:

measuring the pulse wave signal, the electrocardiosignal and the arterial blood pressure signal;

and setting time limit, amplitude peak value size limit and peak time interval limit, and screening the pulse wave signals, the electrocardiosignals and the arterial blood pressure signals.

Further, the filtering the pulse wave signal and the cardiac signal includes:

filtering the pulse wave signal and the cardiac signal using a discrete wavelet transform. Setting db8 mother wavelets, wherein the decomposition layer number of the pulse wave signals is 8, decomposing approximate coefficients and detail coefficients, zeroing the 8 th level of the approximate coefficients and zeroing the first level of the detail coefficients, and performing soft threshold denoising and coefficient reconstruction on the pulse wave signals;

setting the number of decomposition layers of the electrocardiosignals to be 7, decomposing an approximate coefficient and a detail coefficient, zeroing the 7 th level of the approximate coefficient and zeroing the first level of the detail coefficient, and performing soft threshold denoising and coefficient reconstruction on the electrocardiosignals;

thereby filtering the components of the pulse wave signals which are less than 0.25Hz and more than 31.125Hz, and filtering the components of the electrocardiosignals which are less than 0.5Hz and more than 31.125 Hz.

Further, the extracting systolic pressure, diastolic pressure and mean arterial pressure from the arterial blood pressure signal includes:

acquiring a wave peak value of the arterial blood pressure signal as the systolic pressure;

acquiring a trough value of the arterial blood pressure signal as the diastolic pressure;

obtaining a weighted average of the systolic pressure and the diastolic pressure as the mean arterial pressure; wherein the weight of the systolic pressure is 1, and the diastolic pressure is 2.

Further, the training method for predicting the blood pressure by the deep neural network further comprises the following steps:

before the pulse wave signals and the electrocardiosignals are input into the deep neural network model, normalization processing is further carried out on the pulse wave signals and the electrocardiosignals.

Further, the deep neural network model is an MS-CNN network, the MS-CNN network comprises an input layer, a convolutional layer, a BN layer, a Relu layer, a pooling layer and a full connection layer, the MS-CNN network comprises a plurality of channels, and convolutional kernels of different channels are different in size so as to extract features with different scales.

Further, the inputting the pulse wave signal and the electrocardiosignal into the deep neural network model, extracting different scale features, and extracting multi-scale features from the deep neural network model for multi-task regression prediction includes:

the deep neural network model performs convolution processing and maximum pooling processing on the pulse wave signals and the electrocardiosignals for one time;

extracting different scale characteristics of the pulse wave signals and the electrocardiosignals by each channel in the deep neural network model;

after the features extracted by each channel pass through the BN layer and the Relu layer, the features extracted by each channel are subjected to average pooling by the pooling layer;

and performing regression analysis on the features subjected to average pooling by the full connection layer to obtain the multi-task actual output of the deep neural network model.

Further, the loss function is an MSE loss function.

In another aspect, the present invention further includes a computer device including a memory and a processor, where the memory is used to store at least one program, and the processor is used to load the at least one program to perform the training method for predicting blood pressure by a deep neural network in the embodiment.

In another aspect, the present invention further includes a storage medium in which a program executable by a processor is stored, and the program executable by the processor is used for executing the training method for predicting blood pressure by a deep neural network in the embodiment.

The invention has the beneficial effects that: according to the training method for predicting the blood pressure by the deep neural network, the deep neural network model obtained by training has the capability of predicting the blood pressure according to the pulse wave signals and the electrocardiosignals with high precision, online or offline blood pressure measurement can be performed based on the pulse wave signals and the electrocardiosignals of the human body, and compared with the existing blood pressure measurement method, the trained deep neural network model has the advantages of continuity, real time, convenience in operation and the like. The deep neural network model obtained by training the deep neural network blood pressure prediction training method in the embodiment can be used for non-therapeutic purposes, for example, the deep neural network model is used for correcting a blood pressure measuring instrument, a correction process does not need to depend on a special instrument, only a tester is required to operate the trained deep neural network model together with a computer to implement correction, correction conditions are provided for a user, the user can conveniently correct the blood pressure measuring instrument frequently, and therefore the use effect of the blood pressure measuring instrument is improved.

Drawings

FIG. 1 is a flowchart of a training method for predicting blood pressure by a deep neural network according to an embodiment;

FIG. 2 is a schematic diagram of a training method for predicting blood pressure by a deep neural network in an embodiment;

fig. 3 is a structural diagram of an MS-CNN network in an embodiment.

Detailed Description

In this embodiment, referring to fig. 1, the training method for predicting blood pressure by using a deep neural network includes the following steps:

s1, acquiring a pulse wave signal, an electrocardiosignal and an arterial blood pressure signal;

s2, filtering the pulse wave signals and the electrocardiosignals;

s3, extracting systolic pressure, diastolic pressure and average arterial pressure from the arterial blood pressure signal;

s4, inputting the pulse wave signals and the electrocardiosignals into a deep neural network model, extracting features with different scales by the deep neural network model, and performing multi-task regression prediction;

s5, taking systolic pressure, diastolic pressure and average arterial pressure as expected output of the deep neural network model, and determining a value of a loss function according to actual output and the expected output of the deep neural network model;

and S6, when the value of the loss function meets the convergence condition, finishing the training of the deep neural network model.

The principle of steps S1-S6 is shown in fig. 2, where PPG represents a pulse wave signal, ECG represents an electrocardiographic signal, SBP represents a systolic pressure, DBP represents a diastolic pressure, and MAP represents a mean arterial pressure.

In this embodiment, the deep neural network model to be trained is an MS-CNN network. Referring to fig. 3, the MS-CNN network includes an input layer and a maximum pooling layer, and then is divided into 3 channels, each channel includes a plurality of convolution layers, a BN layer, and a Relu layer, respectively, and convolution kernels of different channels are different in size from each other so as to extract different scale features; after 3 channels, there are successively a pooling layer and a full-link layer.

In step S1, a pulse wave signal PPG, an electrocardiographic signal ECG, and an arterial blood pressure signal ABP are acquired to construct a training set and a test set. Wherein, can measure multistage pulse wave signal, multistage electrocardiosignal and multistage arterial blood pressure signal directly from the human body. Next, a preprocessing process of the pulse wave signal, the electrocardiographic signal, and the arterial blood pressure signal is performed. Specifically, the measured pulse wave signals, the electrocardiosignals and the arterial blood pressure signals are screened, the specific screening standards comprise duration limitation, amplitude peak value size limitation and peak time interval limitation, for example, if the duration limitation is set to be more than 8 minutes, the pulse wave signals, the electrocardiosignals and the arterial blood pressure signals with the duration more than 8 minutes can be screened; setting the peak time interval to be more than 0.6s, screening out the arterial blood pressure signals with the peak time interval of more than 0.6 s; and the interference of abnormal values and false peak signals can be eliminated according to the limitation of the amplitude peak value, so that the screened pulse wave signals, the electrocardiosignals and the arterial blood pressure signals are effective signals.

Steps S2 and S3 are still the preprocessing process for the pulse wave signal, the electrocardiographic signal, and the arterial blood pressure signal. In step S2, the pulse wave signal and the electrocardiographic signal are filtered by discrete wavelet transform, so as to filter out components smaller than 0.25Hz and larger than 31.125Hz in the pulse wave signal and filter out components smaller than 0.5Hz and larger than 31.125Hz in the electrocardiographic signal. Specifically, by setting the number of decomposition layers of the pulse wave signal to 8, an approximation coefficient and a detail coefficient are decomposed, the 8 th level of the approximation coefficient is set to zero, the first level of the detail coefficient is set to zero, and soft threshold denoising and coefficient reconstruction are performed on the pulse wave signal; by setting the number of decomposition layers of the electrocardiosignals to be 7, an approximation coefficient and a detail coefficient are decomposed, the 7 th level of the approximation coefficient is set to zero, the first level of the detail coefficient is set to zero, and the electrocardiosignals are subjected to soft threshold denoising and coefficient reconstruction.

In step S3, a peak value of the arterial blood pressure signal is obtained as a systolic pressure SBP, a trough value of the arterial blood pressure signal is obtained as a diastolic pressure DBP, and a weighted average of the systolic pressure SBP and the diastolic pressure DBP is obtained as an average arterial pressure MAP; the systolic pressure is weighted to 1, and the diastolic pressure is 2, i.e., the mean arterial pressure MAP is calculated by the formula MAP ═ SBP +2 DBP)/3.

After step S3 is executed, the pulse wave signal and the electrocardiographic signal are normalized before being input to the deep neural network model. Specifically, the normalization is performed using the formula x is a point in the normalized signal segment, x_max、x_minRespectively a maximum point and a minimum point in the signal segment. The pulse wave signals and the electrocardiosignals are normalized, so that the distribution areas of the pulse wave signals and the electrocardiosignals are [ -1,1 [ -1 [ ]]The influence of amplitude difference can be reduced, and the distribution of the network input ends is similar so as to lead the neural network to be converged better.

In step S4, the pulse wave signal and the electrocardiographic signal are input to the deep neural network model, and features of different scales are extracted from the deep neural network model to perform the multi-task regression prediction. Referring to fig. 3, the deep neural network model firstly performs convolution processing and maximum pooling processing on pulse wave signals and electrocardiosignals for one time, then different scale features are extracted from the pulse wave signals and the electrocardiosignals by different receptive fields of channels in the deep neural network model, the features extracted from the channels are averaged and pooled by the pooling layer after the different scale features extracted from the channels pass through a BN layer and a Relu layer in the channels, and the features after the average pooling are subjected to regression analysis by the full connection layer, so that the actual output of the deep neural network model is obtained. Wherein, each channel extracts features through a plurality of convolution layers and can accelerate the convergence speed through the BN layer; the features extracted from each channel are processed by the Relu layer, so that the gradient can be prevented from disappearing, deep training is facilitated, and the over-fitting problem can be relieved; average pooling is performed on 3 channels respectively, which can reduce feature redundancy.

In step S5, the actual output of the deep neural network model is used as a predicted value, and error calculation is performed on the systolic pressure, the diastolic pressure, and the mean arterial pressure which are used as expected outputs, specifically, according to the actual output and the expected output of the deep neural network model, an MSE loss function is used for calculation to obtain a value of the loss function, if the value of the loss function meets a convergence condition, for example, the value of the loss function is less than a preset value, it can be considered that the training on the deep neural network model has reached the standard, and the training on the deep neural network model is ended. Otherwise, that is, the value of the loss function does not satisfy the convergence condition, for example, the value of the loss function is greater than the preset value, the process may return to step S1 to continue to execute the next round of training process. The values of the loss function obtained by each round of training process that has been executed can also be arranged into a loss curve, and whether to finish the training of the deep neural network model can be determined according to the change of the loss curve.

The deep neural network model trained in the steps S1 to S6 has the ability of predicting blood pressure with high accuracy from pulse wave signals and electrocardiographic signals, that is, the pulse wave signals and electrocardiographic signals measured from a human body are input into the deep neural network model for learning, and the deep neural network model can extract blood pressure information contained in the pulse wave signals and electrocardiographic signals, so as to output values such as systolic pressure, diastolic pressure and average pulse pressure.

On the basis of executing the steps S1-S6, a multitask training model for learning the relation between the pulse wave and the electrocardiosignal characteristics and the blood pressure can be designed, the relevance between the pulse wave and the electrocardiosignal characteristics and the blood pressure is analyzed by utilizing a multitask network, the differences among different tasks are considered, the learned characteristics of different tasks are shared, overfitting to specific tasks is reduced, the adaptability of the network model to different tasks is improved, the model is promoted to predict the continuous blood pressure progress, and meanwhile, the learning time is short.

The deep neural network model trained by the training method provided by the embodiment can realize that the mean errors of systolic pressure, diastolic pressure and mean arterial pressure are respectively 0.007, 0.022 and 0.009mmHg, the mean absolute errors are respectively 4.04, 2.29 and 2.46mmHg, the standard deviations are respectively 5.81, 3.55 and 3.58mmHg, the Pearson correlation coefficients are respectively 0.96, 0.92 and 0.94, the AAMI standard is met, and the A evaluation is achieved in the BHS index, so that the method has better feasibility and effectiveness. Compared with the conventional method for predicting blood pressure by using ECG and PPG signals and a machine learning method, the average absolute errors of the obtained systolic pressure prediction accuracy and the diastolic pressure prediction accuracy are respectively 11.17mmHg and 5.35mmHg, and the accuracy of the blood pressure prediction by the method is greatly improved. Compared with the existing method for predicting blood pressure by combining ECG signals with deep learning, the average absolute errors of the prediction accuracy of systolic pressure and diastolic pressure respectively reach 7.10 mmHg and 4.61mmHg, and the accuracy of the method for predicting blood pressure is obviously improved. In general, the method realizes high-precision blood pressure prediction without a calibration process, and provides a feasible method for realizing continuous blood pressure measurement in the wearable device.

The deep neural network model trained through steps S1-S6 may be used for non-therapeutic purposes, such as using the deep neural network model for calibration of a blood pressure measuring instrument. The deep neural network model in the embodiment is used for processing the pulse wave signals and the electrocardiosignals and outputting values of systolic pressure, diastolic pressure, average pulse pressure and the like of the tester. Under the same environmental and physiological conditions, the values of systolic pressure, diastolic pressure, mean pulse pressure, etc. of the test person are considered to be constant in a short time. Because the deep neural network model can extract the blood pressure information contained in the pulse wave signals and the electrocardiosignals, the values of systolic pressure, diastolic pressure, average pulse pressure and the like output by the deep neural network model can be used as standard values, a blood pressure measuring instrument is used for measuring the blood pressure of a tester, the systolic pressure, the diastolic pressure and the average pulse pressure measured by the blood pressure measuring instrument are respectively compared with the systolic pressure, the diastolic pressure and the average pulse pressure output by the deep neural network model, if the systolic pressure measured by the blood pressure measuring instrument is consistent with the systolic pressure output by the deep neural network model or the error is not more than the threshold value, the error between the systolic pressure measured by the blood pressure measuring instrument and the systolic pressure output by the deep neural network model is considered to be in the normal range of the working parameters of the blood pressure measuring instrument without correction, if the error between the systolic pressure measured by the blood pressure measuring instrument and the systolic pressure output by the deep neural network model is more than the threshold value, and then, the blood pressure measuring instrument is considered to need to be corrected, the working parameters of the blood pressure measuring instrument are adjusted, and the systolic pressure of the tester is measured again until the systolic pressure measured by the blood pressure measuring instrument is consistent with the systolic pressure output by the deep neural network model or the error is not greater than the threshold value, so that the correction of the systolic pressure measuring function of the blood pressure measuring instrument is completed. Similarly, correction of the measurement function of the diastolic blood pressure and the average pulse pressure of the blood pressure measuring instrument can be performed.

By using the training method for predicting blood pressure by using the deep neural network in the embodiment, the trained deep neural network model has the capability of identifying corresponding arterial blood pressure according to the pulse wave signals and the electrocardiosignals, and can perform online or offline blood pressure measurement based on the measured pulse wave signals and the electrocardiosignals of the human body. The deep neural network model obtained by training the deep neural network blood pressure prediction training method in the embodiment can be used for non-therapeutic purposes, for example, the deep neural network model is used for correcting a blood pressure measuring instrument, a correction process does not need to depend on a special instrument, only a tester is required to operate the trained deep neural network model together with a computer to implement correction, correction conditions are provided for a user, the user can conveniently correct the blood pressure measuring instrument frequently, and therefore the use effect of the blood pressure measuring instrument is improved.

The training method for predicting the blood pressure by the deep neural network in the embodiment can be implemented by writing a computer program for implementing the training method for predicting the blood pressure by the deep neural network in the embodiment, writing the computer program into a computer device or a storage medium, and when the computer program is read out and operated, implementing the training method for predicting the blood pressure by the deep neural network in the embodiment, thereby achieving the same technical effect as the reminding system for preventing fatigue driving in the embodiment.

It should be noted that, unless otherwise specified, when a feature is referred to as being "fixed" or "connected" to another feature, it may be directly fixed or connected to the other feature or indirectly fixed or connected to the other feature. Furthermore, the descriptions of upper, lower, left, right, etc. used in the present disclosure are only relative to the mutual positional relationship of the constituent parts of the present disclosure in the drawings. As used in this disclosure, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. In addition, unless defined otherwise, all technical and scientific terms used in this example have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description of the embodiments herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this embodiment, the term "and/or" includes any combination of one or more of the associated listed items.

It will be understood that, although the terms first, second, third, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element of the same type from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples, or exemplary language ("e.g.," such as "or the like") provided with this embodiment is intended merely to better illuminate embodiments of the invention and does not pose a limitation on the scope of the invention unless otherwise claimed.

It should be recognized that embodiments of the present invention can be realized and implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The methods may be implemented in a computer program using standard programming techniques, including a non-transitory computer-readable storage medium configured with the computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner, according to the methods and figures described in the detailed description. Each program may be implemented in a high level procedural or object oriented programming language to communicate with a computer system. However, the program(s) can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.

Further, operations of processes described in this embodiment can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described in this embodiment (or variations and/or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) collectively executed on one or more processors, by hardware, or combinations thereof. The computer program includes a plurality of instructions executable by one or more processors.

Further, the method may be implemented in any type of computing platform operatively connected to a suitable interface, including but not limited to a personal computer, mini computer, mainframe, workstation, networked or distributed computing environment, separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, and the like. Aspects of the invention may be embodied in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, optically read and/or write storage medium, RAM, ROM, or the like, such that it may be read by a programmable computer, which when read by the storage medium or device, is operative to configure and operate the computer to perform the procedures described herein. Further, the machine-readable code, or portions thereof, may be transmitted over a wired or wireless network. The invention described in this embodiment includes these and other different types of non-transitory computer-readable storage media when such media include instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor. The invention also includes the computer itself when programmed according to the methods and techniques described herein.

A computer program can be applied to input data to perform the functions described in the present embodiment to convert the input data to generate output data that is stored to a non-volatile memory. The output information may also be applied to one or more output devices, such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including particular visual depictions of physical and tangible objects produced on a display.

The above description is only a preferred embodiment of the present invention, and the present invention is not limited to the above embodiment, and any modifications, equivalent substitutions, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention as long as the technical effects of the present invention are achieved by the same means. The invention is capable of other modifications and variations in its technical solution and/or its implementation, within the scope of protection of the invention.