Mental illness screening system, chip, apparatus, and computer-readable storage medium

Abstract

The present invention relates to the technical field of medical examination, and provides a mental illness screening system, a chip, an apparatus, and a computer-readable storage medium. The system for screening for mental illness comprises the following modules: an image acquisition module, an ROI detection module, an ROI extraction module, a cropping module, a standardization module, an erythema segmentation module, a statistical feature extraction module, and a report generation module. By using the system of the present invention, it is possible to detect or diagnose whether an individual is suffering from depression. The method for using the system for screening for mental illness of the present invention is simple and highly accurate, and can quickly and conveniently complete a diagnosis procedure.

Description

A system, chip, device and computer readable storage medium for screening mental illness TECHNICAL FIELD

[0001] The present application relates to the technical field of medical detection, and in particular relates to a system, chip, device and computer readable storage medium for screening mental illness. BACKGROUND

[0002] Diagnosis of depression usually relies on assessment of clinical symptoms, which needs to be performed in a hospital, and the process is relatively complex and subjective.

[0003] The article Identification of adolescent patients with depression via assessment of the niacin skin flushing response. Journal of Affective Disorders by Wang, J et al. discloses that the skin niacin flushing response (NSFR) has specific performance in depression, specifically, the NSFR of adolescent patients with depression is significantly different from that of healthy controls, and the NSFR has the potential to be used as a biomarker for depression. However, the article does not disclose how the NSFR is used in the system diagnosis of depression, and cannot be directly applied to the technical field of medical detection.

[0004] There is a lack of a screening system that is direct, objective, fast and accurate for diagnosis of depression in the prior art. SUMMARY

[0005] The present application provides a system, chip, device and computer readable storage medium for screening mental illness, which addresses the deficiencies of the prior art.

[0006] The specific technical solutions of the present application are as follows:

[0007] In a first aspect, the present application provides a system for screening mental illness, comprising the following modules:

[0008] (1) An image acquisition module: performing a skin niacin flushing response test on a subject, taking a photograph to acquire an image;

[0009] (2) An ROI detection module: processing the image obtained from module (1) to obtain rectangular coordinates of the inner frame of the niacin patch;

[0010] (3) An ROI extraction module: using the rectangular coordinates obtained from module (2) to extract the content in the inner frame of the niacin patch and align it to obtain a standard rectangle;

[0011] (4) Cutting module: taking the erythema reaction red macule as the center, the standard rectangle obtained by module (3) is cut to obtain a segmentation map;

[0012] (5) Standardization module: the segmentation map obtained by module (4) is standardized to obtain a standard segmentation map;

[0013] (6) Erythema segmentation module: the standard segmentation map obtained by module (5) is screened to obtain a feature map;

[0014] (7) Statistical feature extraction module: the area of the erythema region in the feature map is extracted and summed to obtain a detection value;

[0015] (8) Report generation module: output the erythema region trend chart and the detection value;

[0016] Specifically, the mental illness is depression;

[0017] Specifically, the photographing collection method is:

[0018] ①Collect the reference image of the inner side of the forearm of the subject without nicotinic acid, and the time is recorded as t=0;

[0019] ②After 1 minute of using the nicotinic acid patch, the nicotinic acid patch is removed, and the image of the skin reaction is captured every 10 seconds for 10 minutes. The image at time t is recorded as x t ;

[0020] Specifically, the number of nicotinic acid patches is greater than or equal to 1, and is coated with a 60mM nicotinic acid methyl ester aqueous solution.

[0021] Specifically, the skin nicotinic acid flushing reaction is that the skin contacting nicotinic acid will produce a flushing reaction red macule;

[0022] Specifically, the processing method in the ROI detection module is:

[0023] c. Read all x t gray scale images, denoted as x t,gray , remove the redundant boundary information in x t,gray to obtain the processed gray scale image;

[0024] d. Obtain the maximum frame of the processed gray scale image in step a, which corresponds to the inner frame of the nicotinic acid patch and is fitted as a rectangle to obtain the rectangular vertex coordinates;

[0025] Specifically, the removal of redundant boundary information in x t,gray in step a is to use the GaussianBlur function in the cv2 module in Python to perform Gaussian blur on x t,gray ;

[0026] Specifically, the step b said to obtain the maximum frame of the gray image processed in step a is: find its maximum frame through the Canny, FindContours and ContourArea functions in the cv2 module in Python;

[0027] Specifically, the step b said to fit into a rectangle is: use the MinAreaRect and BoxPoints functions in the cv2 module in Python to fit the inner frame of the niacin patch into a rectangle;

[0028] Specifically, as the area of the maximum frame in step b is greater than the preset value 300000, repeat the operation of step a and step b;

[0029] Specifically, the extraction method in the ROI extraction module is: read the color image of x t , denoted as x t,RGB , use the rectangular vertex coordinates obtained in step b and the GetPerspectiveTransform and GetPerspective functions in the cv2 module in Python to extract the content in the inner frame from x t,RGB , and align it to a standard rectangle;

[0030] Specifically, the cutting method in the cutting module is: take each erythema reaction erythema as the center, and cut out a segment of 1.8x2.2cm 2 from the standard rectangle, denoted as x t,i ;

[0031] Specifically, the standardization processing method in the standardization module is: extract the mean value of the non-erythema pixels around x t,i , denoted as mean t,i , and adjust the image brightness based on mean t,i to obtain a standardized image, denoted as x t,i,norm , x t,i,norm =x t,i -mean t,i +150;

[0032] Specifically, the screening method of the erythema segmentation module is:

[0033] S1, convert the pixel points in x t,i,norm into two states: black and white, and generate a segmentation image (1);

[0034] S2, remove small noise from the segmentation image (1) obtained in step S1 to obtain a segmentation image (2);

[0035] S3, remove small holes from the segmentation map (2) obtained in step S2 to obtain a segmentation map (3);

[0036] S4, screen out small components from the segmentation map (3) obtained in step S3 to obtain its connected components to obtain a feature map;

[0037] Specifically, the method for converting the pixel points in x t,i,norm into two states in step S1 is as follows: obtaining diff t,i,norm by subtracting x 0,i,norm from x t,i , calculating the standard deviation std t,i of diff t,i , and binarizing diff t,i with the standard deviation as a threshold.

[0038] Specifically, the x 0,i,norm is a normalized image at the 0th moment.

[0039] Specifically, the method for removing small noise in step S2 is as follows: using the MedianBluer function in the cv2 module in Python to perform median filtering on the segmentation map (1) obtained in step S1.

[0040] Specifically, the method for removing small holes in step S3 is as follows: using the MorphologyEx function in the cv2 module in Python to perform morphological dilation on the segmentation map (2) obtained in step S2.

[0041] Specifically, the method for screening out small components in step S4 is as follows: using the ConnectedComponentsWithStats function in the cv2 module in Python on the segmentation map (3).

[0042] Specifically, the method for extracting the area of the erythema region in the feature map in the statistical feature extraction module is as follows: extracting the area of the erythema region in each feature map.

[0043] Specifically, the summation in the statistical feature extraction module is as follows: summing up the areas of the erythema regions in the obtained feature maps at each moment to obtain a detection value.

[0044] Specifically, the detection value >= 4000 indicates a tendency of depression.

[0045] The detection value < 4000 indicates mental health.

[0046] In a second aspect, the present application also provides a chip comprising a processor and a memory,

[0047] Specifically, the memory stores machine executable instructions of the above system.

[0048] In particular, the processor can execute machine executable instructions in the memory to run the above-mentioned system.

[0049] In a third aspect, the application further provides a device for screening mental illness, characterized by comprising a display, a controller and the above-mentioned chip.

[0050] In particular, the display is used to display the operation state and the image.

[0051] In particular, the controller is used to control the operation of the device.

[0052] In a fourth aspect, the application further provides a computer readable storage medium storing a program for executing the above-mentioned system.

[0053] The application has the following advantages:

[0054] (1) The application combines image processing technology and statistical methods to accurately analyze the degree of skin reaction to niacin;

[0055] (2) By testing the patient with niacin and taking pictures of the skin reaction, the algorithm converts the image data into quantifiable biomarker indicators, effectively diagnosing depression.

[0056] (3) The application is characterized by its high objectivity and accuracy, providing a new technical means for the diagnosis of depression compared to traditional symptom evaluation methods. BRIEF DESCRIPTION OF DRAWINGS

[0057] The application will be described in further detail below in conjunction with the accompanying drawings and preferred embodiments, but those skilled in the art will appreciate that the drawings are only drawn for the purpose of explaining the preferred embodiments and therefore should not be considered as limiting the scope of the application. In addition, unless specifically indicated, the drawings only schematically represent the composition or structure of the described objects and can contain exaggerated displays, and the drawings are not necessarily drawn to scale.

[0058] Figure 1 is an image taken after 1 minute of niacin flushing reaction;

[0059] Figure 2 shows from top to bottom: a more obvious niacin flushing reaction image, a feature map of the example processing of the more obvious niacin flushing reaction image;

[0060] Figure 3 shows from top to bottom: a less obvious niacin flushing reaction image, a feature map of the example processing of the less obvious niacin flushing reaction image. DETAILED DESCRIPTION

[0061] The following non-limiting examples can make the ordinary skilled in the art more comprehensive understanding of the present application, but not in any way limit the present application. The following content is only an exemplary description of the scope of the present application, those skilled in the art can make many changes and modifications to the application of the present application according to the disclosed content, and it should also belong to the scope of the present application claimed.

[0062] The technical details of the present application are described in detail by the following examples. It should be noted that the examples are only used to further illustrate the technical features of the present application, and are not intended to limit the scope of the present application.

[0063] Unless otherwise specified, all goods or reagents in the present application are purchased through market channels.

[0064] The present application will be described in further detail below in conjunction with specific examples, but the embodiments of the present application are not limited thereto.

[0065] The system for screening mental illness provided by the embodiment of the present application, wherein the mental illness is depression, and the system comprises the following modules:

[0066] (1) Image acquisition module: performing skin niacin flare response test on the subject, taking pictures to collect images.

[0067] First, the reference image of the inner side of the forearm of the subject without niacin is collected, and the time is recorded as t = 0; after the subject uses a 60 mM niacin patch on the inner side of the forearm for 1 min, the niacin patch is removed, and the image of the skin response is captured every 10 seconds, lasting for 10 minutes, and the image at time t is recorded as x t .

[0068] In actual application, the present application collects at time 0, which corresponds to the reference image without niacin. If the reference image is set as the first image taken after 1 min of reaction, it is unreasonable for the case of faster reaction, because there may be erythema at this time, as shown in FIG. 1.

[0069] (2) ROI detection module: processing the image obtained from module (1) to obtain the rectangular coordinates of the inner frame of the niacin patch.

[0070] In actual application, first, for the image at time t, read its gray image and record it as x t,gray , use the GaussianBlur function in the cv2 module in Python to perform Gaussian blur on it to obtain the processed gray image. The purpose of this is to remove redundant boundary information.

[0071] The maximum border in the processed gray image is found by Canny, findContours and contourArea functions in the cv2 module in Python, which corresponds to the inner border of the niacin patch.

[0072] If the area of the border is greater than the preset value (300000) during processing, repeat the operation of Gaussian blur and finding the maximum border. The purpose of this is to ensure that the found niacin patch border corresponds to the inner border of the niacin patch rather than the outer border of the niacin patch.

[0073] The inner border of the niacin patch is fitted with a rectangle using the minAreaRect and boxPoints functions in the cv2 module in Python, and the coordinates of the four vertices of the rectangle are obtained.

[0074] (3) ROI extraction module: use the rectangular coordinates obtained from module (2) to extract and obtain the content in the inner border of the niacin patch and align it to a standard rectangle.

[0075] In practical applications, first read the color image, denoted as x t,RGB , use the obtained rectangular vertex coordinates and the GetPerspectiveTransform and GetPerspective functions in the cv2 module in Python to extract the content in the inner border from x t,RGB , and align it to a standard rectangle.

[0076] (4) Cutting module: take the erythema reaction erythema as the center and cut the standard rectangle obtained from module (3) to obtain a segmentation map.

[0077] In practical applications, when cutting, take the erythema reaction erythema as the center and cut the standard rectangle to obtain a segment with a size of 1.8 x 2.2 cm2, denoted as x t,i .

[0078] (5) Standardization module: standardize the segmentation map to obtain a standard segmentation map.

[0079] In practical applications, extract the mean value of the non-erythema pixels around x t,i , denoted as mean t,i , and adjust the image brightness based on it to obtain a standardized image. The standardized image obtained at time t is denoted as x t,i,norm ,x t,i,norm = x t,i -mean t,i +150.

[0080] (6) Erythema segmentation module: screen the standard segmentation map to obtain a feature map.

[0081] In practical applications, for a standardized image x at time t≠0 t,i,norm The erythema segmentation involves the following steps. First, x... t,i,norm With x 0,i,norm Difference to obtain diff t,i Calculate diff t,i standard deviation std t,i And use this as a threshold to binarize diff t,i The purpose of doing this is to make x t,i,norm The pixels in the image are converted into two states: black and white, making our segmentation algorithm more robust to uncontrollable factors such as changes in ambient light. Thus, the segmentation image (1) is obtained.

[0082] The medianBluer function in the cv2 module of Python is used to perform median filtering on the segmentation image (1). The purpose of this is to remove small noise and make the result more accurate. Thus, the segmentation image (2) is obtained.

[0083] The segmentation image (2) is morphologically expanded using the morphologyEx function in the cv2 module of Python. The purpose of this is to remove small holes in the segmentation image to make the result more accurate. Thus, the segmentation image (3) is obtained.

[0084] Finally, the connectedComponentsWithStats function in the cv2 module of Python is used to obtain the connected components in the segmentation map (3) and filter out small components to make the results more reliable, thus obtaining the feature map.

[0085] The ROI detection module, ROI extraction module, cropping module, standardization module and erythema segmentation module of this invention were used to process the images of more obvious nicotinic acid erythema in Figure 2 and the images of less obvious nicotinic acid erythema in Figure 3. The results are shown in Figures 2 and 3.

[0086] As shown in Figure 2, as the erythema of the flushing reaction becomes lighter and more irregular, the feature map obtained by the technical solution of the present invention still appears complete and connected.

[0087] Figure 3 shows a skin image of a mild niacin flushing reaction. The technical solution of the present invention can still extract a clear and complete feature map.

[0088] This invention reduces the impact of uncontrollable factors such as changes in ambient light, removes small noises and voids, and eliminates small components, making the processing results more reliable.

[0089] The mental disease screening system can accurately analyze the degree of skin reaction to niacin, does not miss small differences, and has small errors; can convert image data into quantifiable biomarker indicators, effectively diagnoses depression, and has high objectivity and accuracy.

[0090] The statistical feature extraction module extracts the area of the erythema region in the feature map, sums up, and obtains a detection value.

[0091] The report generation module: based on the experimental data under the preliminary experiment, the detection value >=4000 is a depression tendency, and the detection value <4000 is a mental health.

[0092] Based on the same inventive concept, the embodiment of the present application also provides a chip based on a mental disease screening system, including a processor and a memory.

[0093] The memory stores machine executable instructions of the following modules:

[0094] (1) Image acquisition module: performing skin niacin flushing reaction test on the subject, photographing and collecting, and obtaining images;

[0095] (2) ROI detection module: processing the images obtained by module (1) to obtain the rectangular coordinates of the inner frame of the niacin patch;

[0096] (3) ROI extraction module: extracting using the rectangular coordinates obtained by module (2) to obtain the content in the inner frame of the niacin patch and aligning to obtain a standard rectangle;

[0097] (4) Cutting module: taking the flushing reaction erythema as the center, cutting the standard rectangle obtained by module (3) to obtain a segmentation image;

[0098] (5) Standardization module: standardizing the segmentation image obtained by module (4) to obtain a standard segmentation image;

[0099] (6) Erythema segmentation module: screening the standard segmentation image obtained by module (5) to obtain a feature map;

[0100] (7) Statistical feature extraction module: extracting the area of the erythema region in the feature map, summing up, and obtaining a detection value;

[0101] (8) Report generation module: outputting the erythema region trend chart and the detection value.

[0102] The processor executes the machine executable instructions of the above modules in the memory.

[0103] The memory can include a volatile memory such as a random access memory, and can also include a non-volatile memory.

[0104] In some embodiments, the application selects a non-volatile memory.

[0105] The processor can be a general purpose processor, a special purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array, or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc.

[0106] In some embodiments, the application selects a general purpose processor.

[0107] Based on the same inventive concept, the application also provides a device for screening mental illness, comprising a display, a controller and a chip based on the system for screening mental illness.

[0108] The display is used to display the operation state and the captured image and the processed image, and the controller is used to control the operation of the device.

[0109] Based on the same inventive concept, the application also provides a computer readable storage medium. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method steps of any of the above systems for screening mental illness.

[0110] The computer readable storage medium can be a non-volatile memory, such as at least one disk memory, and can also be at least one storage device located away from the processor. The processor can be a general purpose processor, a special purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array, or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc.

[0111] In some embodiments, the application selects a non-volatile memory as the memory and a general purpose processor as the processor.

[0112] Although the application has been described herein in relation to the embodiments thereof, other variations and modifications can be understood and effected by those skilled in the art upon reading the drawings and the disclosure. The above description is further detailed in relation to specific preferred embodiments of the application, and should not be construed as limiting the application to only these descriptions. For those skilled in the art, without departing from the concept of the application, a number of simple derivations or substitutions can be made, which should be considered as falling within the scope of protection of the application.

Claims

1. A system for screening of mental illness, characterized by, The method comprises the following modules: (1) image acquisition module: performing skin nicotinic acid flushing reaction test on the subject, taking pictures to acquire images; (2) ROI detection module: processing the images obtained in module (1) to obtain rectangular coordinates of the inner frame of the nicotinic acid patch; (3) ROI extraction module: extracting using the rectangular coordinates obtained in module (2) to obtain the content in the inner frame of the nicotinic acid patch and aligning to obtain a standard rectangle; (4) cutting module: taking the flushing reaction erythema as the center, cutting the standard rectangle obtained in module (3) to obtain a segmentation image; (5) standardization module: standardizing the segmentation image obtained in module (4) to obtain a standard segmentation image; (6) erythema segmentation module: screening the standard segmentation image obtained in module (5) to obtain a feature image; (7) statistical feature extraction module: extracting the area of the erythema region in the feature image, summing up to obtain a detection value; (8) report generation module: outputting an erythema region trend chart and the detection value.

2. The system of claim 1, wherein, The mental illness is depression.

3. The system of claim 1, wherein, The photographing acquisition method is: ①acquiring a reference image of the inner side of the forearm of the subject without nicotinic acid, and the time is recorded as t=0; (ii) After 1 minute of use of the nicotinic acid patch by the subject, the nicotinic acid patch is removed, and images of the skin reaction are captured every 10 seconds for 10 minutes, and the image for time t is denoted as x t .

4. The system of claim 3, wherein, The number of the nicotinic acid patches is greater than or equal to 1, and the nicotinic acid patches are coated with a 60 mM nicotinic acid methyl ester aqueous solution.

5. The system of claim 1, wherein, The skin nicotinic acid flushing reaction is that the skin contacting with nicotinic acid will produce flushing reaction erythema.

6. The system of claim 3, wherein, The processing method in the ROI detection module is: a. Read all x t grayscale maps, denoted as x t,gray , remove redundant boundary information in x t,gray to obtain processed grayscale maps; b. obtaining the maximum frame of the gray image processed in step a, which corresponds to the inner frame of the nicotinic acid patch and is fitted as a rectangle to obtain rectangular vertex coordinates.

7. The system of claim 6, wherein, The removal of x in step a t,gray The boundary information of the redundancy is: using the GaussianBlur function in the cv2 module in Python to perform Gaussian blur on x t,gray ​ In step b, the maximum frame of the gray image processed in step a is obtained by using the Canny, FindContours and ContourArea functions in the cv2 module in Python; In step b, the inner frame of the nicotinic acid patch is fitted as a rectangle by using the MinAreaRect and BoxPoints functions in the cv2 module in Python.

8. The system of claim 7, wherein, If the area of the maximum frame as described in step b is greater than a preset value of 300000, the steps a and b are repeated.

9. The system of claim 8, wherein, The method of extraction in the ROI extraction module is: reading the color image x t , denoted as x t,RGB , using the rectangular vertex coordinates obtained in step b and the GetPerspectiveTransform and GetPerspective functions in the cv2 module in Python to extract the content in the inner border from x t,RGB , and aligning to a standard rectangle.

10. The system of claim 9, wherein, The method of cutting in the cutting module is: taking the erythema as the center, cutting a standard rectangle with a size of 1.8 x 2.2 cm 2 segment, denoted as x t,i .

11. The system of claim 10, wherein, The standardization process described in the standardization module is as follows: for x t,i Extract the mean value of the surrounding non-red patch pixels, denoted as mean. t,i and with mean t,i To uniformly adjust the image brightness and obtain a standardized image, the standardized image obtained at time t is denoted as x. t,i,norm x t,i,norm =x t,i -mean t,i +150.

12. The system of claim 11, wherein, The screening method in the erythema segmentation module is: S1, convert the pixel points in x t,i,norm to two states: black and white, to generate a segmentation map (1); S2, removing small noise from the segmentation image (1) obtained in step S1 to obtain a segmentation image (2); S3, removing small holes from the segmentation image (2) obtained in step S2 to obtain a segmentation image (3); S4, screening small components from the segmentation image (3) obtained in step S3 to obtain its connected components to obtain a feature image.

13. The system of claim 12, wherein, The step S1 described above is to x t,i,norm The method for converting pixels in a given area to two states is as follows: x t,i,norm With x 0,i,norm Difference to obtain diff t,i Calculate diff t,i standard deviation std t,i And use this as a threshold to binarize diff t,i ; The x 0,i,norm is the normalized image at time 0; In step S2, the method for removing small noise is to perform median filtering on the segmentation image (1) obtained in step S1 by using the MedianBluer function in the cv2 module in Python; In step S3, the method for removing small holes is to perform morphological dilation on the segmentation image (2) obtained in step S2 by using the MorphologyEx function in the cv2 module in Python; The method for screening out small components in step S4 is: using the ConnectedComponentsWithStats function in the cv2 module in Python on the segmentation graph (3).

14. The system of claim 13, wherein, The area of the erythema region in the feature map is extracted in the statistical feature extraction module. The summation in the statistical feature extraction module is: summing up the area of the erythema region of the feature map at each time to obtain a detection value.

15. The system of claim 14, wherein, The detection value >= 4000 is a tendency to depression. The detection value < 4000 is mental health.

16. A chip, characterized by It comprises a processor and a memory, wherein, The memory stores machine executable instructions of the system of any one of claims 1-15; The processor executes the machine executable instructions in the memory to run the system of any one of claims 1-15.

17. A device for screening for a mental disease, characterized by It comprises a display, a controller and the chip of claim 16; The display is: displaying the operating state and the image; The controller is: controlling the device operated by the apparatus.

18. A computer-readable storage medium, characterized in that, It stores a program for executing the system of any one of claims 1-15.

Images