CN112869744B - Auxiliary diagnosis method, system and storage medium for schizophrenia - Google Patents

Abstract

The invention provides a method, a system and a storage medium for auxiliary diagnosis of schizophrenia, which collect eye movement data based on free view experiments, have the characteristics of simplicity, easiness in operation and small manual work load, and provide five eye movement characteristics with interpretability based on eye movement tracking data, so that a machine learning method is successfully introduced into the field of diagnosis of schizophrenia, and objective quantitative indexes and prediction results based on a classifier are provided for doctors to diagnose schizophrenia.

Description

Schizophrenia auxiliary diagnosis method, system and storage medium

Technical Field

The present invention relates to the technical field of mental illness diagnosis, and in particular to a schizophrenia auxiliary diagnosis method, system and storage medium.

Background technique

Schizophrenia is an extremely complex clinical syndrome associated with multiple genes and genetic traits, with the characteristics of early onset, long course, severe functional impairment, and great social harm. At present, the diagnosis of this disease mainly relies on clinical interviews by doctors, lacking good biological indicators, resulting in a high misdiagnosis rate and delayed treatment for patients.

In recent years, eye tracking technology based on iris reflection and video image capture has become increasingly mature, and the accuracy of eye movement data and the convenience of collection have gradually increased. Studies on the brain mechanism of eye movement have shown that the complete eye movement process requires both sensorimotor conversion function and cognitive function guidance. The sensorimotor conversion function converts the visual signals received by the retina into eye movements, while more advanced eye movement processes such as inhibitory eye movement control and prediction of target movement speed require the brain to activate cognitive functions for regulation. Therefore, eye movement indicators can be used as behavioral measurement indicators to explore the human cerebral cortex and subcortical related advanced cognitive processes.

Summary of the invention

The purpose of the present invention is to propose a series of eye movement indicators with significant abnormalities for the abnormal eye movement behavior patterns of existing schizophrenia, and use these indicators as features to predict the risk of schizophrenia in the subjects using machine learning methods.

In order to achieve the above object, the present invention provides a method for auxiliary diagnosis of schizophrenia, characterized in that the specific steps are as follows:

S1, conduct free view experiments on volunteers and patients, collect eye tracking data, and establish a database;

S2, establish a sample feature matrix based on the feature calculation formula;

S3, training support vector machine classification model;

S4, conduct a free view experiment on the subjects, collect eye tracking data, and calculate the sample feature matrix;

S5, input the sample feature matrix of the subject into the classification model trained in step S3, calculate the disease risk according to the classification result, and output the diagnosis conclusion.

Furthermore, the step S1 further includes:

S101, using an eye tracking device to record eye tracking data of a subject, wherein the eye tracking data includes eye gaze position and saccadic path;

S102, establishing a relational database using the subject's serial number and the picture name as primary keys, wherein each record in the database as a sample includes eye tracking data corresponding to the primary key.

Furthermore, the step S2 further includes:

S201, calculate the average visual saccade velocity The calculation process is:

Where n is the number of saccades, Amp _i is the intensity of the saccade of the i-th saccade, and _ti is the duration of the i-th saccade. The speed of the saccade is obtained by dividing the two, and then the average speed of n saccades is calculated.

Furthermore, the step S2 further includes:

S202, calculating the pupil dynamic range, the calculation process is:

Where d _max and d _min represent the maximum pupil diameter and the minimum pupil diameter, respectively. Represents the average pupil diameter.

Furthermore, the step S2 further includes:

S203, calculating the maximum mean pupil ratio MMRPS, the calculation process is:

Furthermore, the step S2 further includes:

S204, calculate the fixation point skewness SF, the calculation process is:

Among them, _x0 and _y0 are the horizontal and vertical coordinates of the geometric center of the image, _xi and _yi represent the horizontal and vertical coordinates of the i-th fixation point, and n is the number of fixation points.

Furthermore, the step S2 further includes:

S205, calculating the effective observation time, the calculation process is:

Among them, n and m represent the number of fixations and the number of saccades, respectively. represents the fixation duration of the ith fixation point, /> represents the saccade duration of the j-th saccade, and the calculated result is the effective observation time, denoted as t _valid .

Furthermore, the step S2 further includes:

S206, merging the calculated five eigenvalues with the corresponding subject serial numbers and picture names to form a sample feature matrix, wherein the dimension of the sample feature matrix is N×5, where N is the total number of samples.

Furthermore, the step S3 further includes:

S301, using samples of patients and volunteers in the database as positive samples and negative samples respectively, to train a support vector machine model;

S302, selecting a Gaussian radial basis function kernel, mapping the low-dimensional features to a high-dimensional feature space, and calculating the distance between the two samples;

S303, using a sequential minimization optimization method to train the support vector machine, and setting the loss function as an average error function.

Furthermore, the calculation process of the distance between the samples is:

Among them, x and x' represent the feature vectors of two samples respectively, and σ is the scaling factor of the kernel function.

Furthermore, in step S5, the calculation process of the diagnosis result includes:

S501, calculate the prediction result according to the following formula:

Among them, p is defined as the number of samples classified as positive in the classification results divided by the total number of samples, and Th is a pre-set threshold;

S502, calculate confidence:

Conf(p)＝2*[sigmoid(10*|p-0.5|)-0.5]

Among them, sigmoid is the activation function, defined as:

On the other hand, the present invention also provides a schizophrenia auxiliary diagnosis system, comprising:

The data acquisition module conducts free-view experiments on volunteers and patients, collects eye tracking data, and builds a database;

Feature extraction module, which establishes a sample feature matrix based on the feature calculation formula;

Model building module, training support vector machine classification model;

The diagnosis output module inputs the sample feature matrix of the subject into the trained classification model, calculates the risk of disease based on the classification results, and outputs the diagnosis conclusion.

On the other hand, the present invention further provides a computer-readable storage medium storing a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the steps of the method as claimed in claim 1 .

The present invention has the following beneficial effects:

The present invention collects eye movement data based on a free view experiment, which is simple and easy to operate and requires little manual work. In addition, based on the eye tracking data, five interpretable eye movement features are proposed, which successfully introduces machine learning methods into the field of schizophrenia diagnosis, and provides doctors with objective quantitative indicators and classifier-based prediction results for diagnosing schizophrenia.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

FIG1 is a flow chart of a method for auxiliary diagnosis of schizophrenia according to the present invention.

FIG. 2 is a schematic diagram of a confidence function of a prediction result of the present invention.

FIG3 is a system framework diagram of a schizophrenia auxiliary diagnosis system of the present invention.

Detailed ways

The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

Eye movement indicators can be used as behavioral measurement indicators to explore the higher-level cognitive processes related to the human cerebral cortex and subcortex. For schizophrenia patients with cognitive dysfunction, their eye movement abnormalities relative to normal people can be used as biological indicators to characterize their mental depression state. Therefore, the eye movement differences of schizophrenia patients can be detected through a specific experimental paradigm, and a series of eye movement indicators with significant abnormalities can be proposed. These indicators can be used as features to predict the risk of schizophrenia in the subjects using machine learning methods.

In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

FIG1 is a method flow chart of a method for assisting diagnosis of schizophrenia according to an embodiment of the present invention. As shown in FIG1 , a method for assisting diagnosis of schizophrenia according to an embodiment of the present invention comprises the following steps:

S1, conduct free view experiments on volunteers and patients, collect eye tracking data, and establish a database.

Specifically, during the experiment, each volunteer or patient (collectively referred to as subjects) was asked to sit in front of the experimental equipment and freely watch the pictures appearing on the screen. At the same time, professional eye tracking equipment was used to record information such as their eye gaze position and saccadic path.

After the experiment is completed, the eye tracking data is processed and a relational database is established with "subject number-picture name" as the primary key, that is, each "subject number-picture name" and its corresponding eye tracking data are regarded as a sample.

S2, establish a sample feature matrix based on the feature calculation formula.

Specifically, in the step S2, it also includes:

S201, calculate the average visual saccade velocity The calculation process is:

Where n is the number of saccades, Amp _i is the intensity of the saccade of the i-th saccade, and _ti is the duration of the i-th saccade. The speed of the saccade is obtained by dividing the two, and then the average speed of n saccades is calculated.

S202, calculating the pupil dynamic range, the calculation process is:

Where d _max and d _min represent the maximum pupil diameter and the minimum pupil diameter, respectively. Represents the average pupil diameter.

S203, calculating the maximum mean pupil ratio MMRPS, the calculation process is:

S204, calculate the fixation point skewness SF, the calculation process is:

Among them, _x0 and _y0 are the horizontal and vertical coordinates of the geometric center of the image, _xi and _yi represent the horizontal and vertical coordinates of the i-th fixation point, and n is the number of fixation points.

S205, calculating the effective observation time, the calculation process is:

Among them, n and m represent the number of fixations and the number of saccades, respectively. represents the fixation duration of the ith fixation point, /> represents the saccade duration of the j-th saccade, and the calculated result is the effective observation time, denoted as t _valid .

S206, merging the calculated five eigenvalues with the corresponding subject serial numbers and picture names to form a sample feature matrix, wherein the dimension of the sample feature matrix is N×5, where N is the total number of samples.

S3, train the support vector machine classification model.

Specifically, in step S3, it also includes:

S301, using the samples of patients and volunteers in the database as positive samples and negative samples respectively, to train a support vector machine model.

S302, select a Gaussian radial basis function kernel, map the low-dimensional features to a high-dimensional feature space, and calculate the distance between the two samples.

The calculation process of the distance between the samples is as follows:

Among them, x and x' represent the feature vectors of two samples respectively, and σ is the scaling factor of the kernel function.

S303, using a sequential minimization optimization method to train the support vector machine, and setting the loss function as an average error function.

S4, conduct a free view experiment on the subjects, collect eye tracking data, and calculate the sample feature matrix.

S5, input the sample feature matrix of the subject into the classification model trained in step S3, calculate the disease risk according to the classification result, and output the diagnosis conclusion.

Specifically, in step S5, the calculation process of the diagnosis result includes:

S501, calculate the prediction result according to the following formula:

Among them, p is defined as the number of samples classified as positive in the classification results divided by the total number of samples, and Th is a pre-set threshold.

S502, calculate confidence. FIG2 is a schematic diagram of the confidence function of the prediction result of the present invention. As shown in FIG2, the confidence calculation function is:

Conf(p)＝2*[sigmoid(10*|p-0.5|)-0.5]

Among them, sigmoid is the activation function, defined as:

FIG3 is a system framework diagram of a schizophrenia auxiliary diagnosis system of the present invention. As shown in FIG3, a schizophrenia auxiliary diagnosis system of the present invention includes:

Data acquisition module 1: Conduct free view experiments on volunteers and patients, collect eye tracking data, and establish a database.

Feature extraction module 2, establishes a sample feature matrix based on the feature calculation formula.

Model building module 3, training support vector machine classification model.

The diagnosis output module 4 inputs the sample feature matrix of the subject into the trained classification model, calculates the disease risk according to the classification result, and outputs the diagnosis conclusion.

In another embodiment of the present invention, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the steps of the above method.

The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or terminal. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art can change, modify, replace and modify the above embodiments within the scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims (4)

1. A schizophrenia auxiliary diagnosis system, comprising: The data acquisition module conducts free-view experiments on volunteers and patients, collects eye tracking data, and builds a database; The feature extraction module establishes a sample feature matrix based on the feature calculation formula. The sample feature matrix includes the average visual saccade velocity, pupil dynamic range, pupil maximum mean ratio, fixation point skewness SF, and effective observation time. The feature extraction module performs the following method: S201, calculate the average visual saccade velocity , and its calculation process is: in, Indicates the number of saccades, /> represents the saccade strength of the i-th saccade, /> represents the duration of the i-th saccade; S202, calculating the pupil dynamic range, the calculation process is: in, and/> Represent the maximum pupil diameter and the minimum pupil diameter respectively,/> represents the average pupil diameter; S203, calculating the maximum mean pupil ratio MMRPS, the calculation process is: S204, calculate the fixation point skewness SF, the calculation process is: in, and/> are the horizontal and vertical coordinates of the geometric center of the image, respectively. and/> represents the horizontal and vertical coordinates of the i-th fixation point, and n is the number of fixation points; S205, calculating the effective observation time, the calculation process is: Among them, n and m represent the number of fixations and the number of saccades, respectively. represents the fixation duration of the ith fixation point, represents the duration of the jth saccade, and the calculated result is the effective observation time, recorded as/> ; S206, merging the calculated five eigenvalues with the corresponding subject serial numbers and picture names to form a sample feature matrix, the dimension of which is N 5, N is the total number of samples; The diagnosis output module inputs the sample feature matrix of the subject into the trained classification model, calculates the risk of disease based on the classification results, and outputs the diagnosis conclusion; The model building module trains the support vector machine classification model, and the model building module performs the following method: S301, using samples of patients and volunteers in the database as positive samples and negative samples respectively, to train a support vector machine model; S302, selecting a Gaussian radial basis function kernel, mapping the low-dimensional features to a high-dimensional feature space, and calculating the distance between the two samples; S303, using a sequential minimization optimization method to train the support vector machine, and setting the loss function as an average error function. 2. A schizophrenia auxiliary diagnosis system according to claim 1, characterized in that the data acquisition module performs the following method: S101, using an eye tracking device to record eye tracking data of a subject, wherein the eye tracking data includes eye gaze position and saccadic path; S102, establishing a relational database using the subject's serial number and the picture name as primary keys, wherein each record in the database as a sample includes eye tracking data corresponding to the primary key. 3. The schizophrenia auxiliary diagnosis system according to claim 2, characterized in that the distance between the samples is calculated by: Among them, x and x' represent the feature vectors of the two samples respectively, and σ is the scaling factor of the kernel function. 4. The schizophrenia auxiliary diagnosis system according to claim 1, characterized in that the diagnosis output module performs the following method: S501, calculate the prediction result according to the following formula: Among them, p is defined as the number of samples classified as positive in the classification results divided by the total number of samples, and Th is a pre-set threshold; S502, calculate confidence: Among them, sigmoid is the activation function, defined as: .