Multi-diagnosis parameter data processing system for traditional Chinese medicine observation, smelling, questioning and cutting

Abstract

The invention discloses a traditional Chinese medicine observation, smell, question and cut multi-diagnosis parameter data processing system, and the system comprises an acquisition module which is used for collecting multi-diagnosis parameter data of observation, smell, question and cut of a patient; the preprocessing module is used for preprocessing the multi-diagnosis parameter data of observation, smelling, question and cut of the patient to obtain target multi-diagnosis parameter data; the intelligent diagnosis module is used for inputting the target multi-diagnosis parameter data into a pre-trained intelligent diagnosis model for analysis to obtain an intelligent diagnosis result; the result output module is used for outputting an intelligent diagnosis result; data in multiple aspects of observation, smell and questionnaire are integrated and analyzed, intelligent diagnosis results are provided for doctors as references, and the doctors are helped to formulate treatment schemes better.

Description

Technical Field

[0001] The invention relates to the field of data processing, in particular to a data processing system for traditional Chinese medicine multi-diagnosis combined with observation, auscultation, questioning and palpation. Background Art

[0002] The four diagnostic methods of observation, auscultation, questioning and palpation can be used to comprehensively understand the patient's condition. Observation is done by observing the patient's complexion, tongue, etc.; auscultation is done by listening to the patient's voice, breathing, etc.; questioning is done by asking about the patient's medical history, symptoms, etc.; palpation is done by touching the patient's pulse to understand the condition. The four diagnostic methods can comprehensively understand the patient's condition and provide a basis for formulating a nursing plan. In reality, the experience and subjective judgment of TCM doctors in observation, auscultation and palpation may have a certain impact on the results, and there is a certain degree of subjectivity and uncertainty. Therefore, there is an urgent need for a TCM multi-diagnosis information processing system to assist TCM doctors in making diagnoses and provide a basis for accurate syndrome differentiation and treatment in TCM. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems in the above-mentioned technologies to a certain extent. To this end, the purpose of the present invention is to propose a data processing system for TCM multi-diagnosis by observation, listening, questioning and palpation, which integrates and analyzes data from multiple aspects of observation, listening, questioning and palpation, provides doctors with intelligent diagnosis results as a reference, and helps doctors to better formulate treatment plans.

[0004] To achieve the above-mentioned purpose, the embodiment of the present invention proposes a data processing system for TCM multi-diagnosis combined with observation, auscultation, inquiry and palpation, including:

[0005] The acquisition module is used to collect the patient's multi-diagnosis data through observation, auscultation, questioning and palpation;

[0006] A preprocessing module is used to preprocess the multi-diagnosis combined reference data of the patient's inspection, auscultation, questioning and palpation to obtain target multi-diagnosis combined reference data;

[0007] An intelligent diagnosis module is used to input the target multi-diagnosis combined data into a pre-trained intelligent diagnosis model for analysis to obtain an intelligent diagnosis result;

[0008] The result output module is used to output the intelligent diagnosis results.

[0009] Preferably, the acquisition module includes:

[0010] The first acquisition submodule is used to acquire images of the patient based on an image acquisition device to obtain a visual image of the patient;

[0011] The second acquisition submodule is used to collect the patient's voice information based on the voice acquisition device to obtain the patient's voice data;

[0012] The third collection submodule is used to obtain patient data based on a questionnaire;

[0013] The fourth acquisition submodule is used to collect the pulse data of the patient based on the sensor to obtain the patient's pulse data;

[0014] The first determination submodule is used to use the patient's visual image, the patient's smell data, the patient's question data and the patient's palpation data as the patient's visual, smell, question and palpation multi-diagnosis combined reference data.

[0015] Preferably, the preprocessing module comprises:

[0016] A first denoising submodule, configured to denoise the patient's visual image to obtain a denoised visual image;

[0017] A second denoising submodule, used to denoise the patient's heard data to obtain denoised heard data;

[0018] A screening submodule, used for screening the patient's question data to obtain screened question data;

[0019] An enhancer module, used for enhancing the cut data of the patient to obtain enhanced hear data;

[0020] The second determination submodule is used to use the denoised visual image, denoised hearing data, screened patient question data and enhanced hearing data as target multi-diagnosis combined reference data.

[0021] Preferably, the first noise reduction submodule includes:

[0022] First classification unit for:

[0023] Take any patient's visual image as the image to be processed;

[0024] Grayscale the image to be processed to obtain a grayscale image;

[0025] The grayscale image is evenly divided into a number of sub-images to obtain a number of grayscale sub-images;

[0026] Take any grayscale sub-image as the first image;

[0027] Obtaining the grayscale value of each pixel in the first image;

[0028] Based on the Otsu threshold segmentation method, determine the grayscale threshold corresponding to the first image;

[0029] Compare the grayscale value of each pixel in the first image with the grayscale threshold corresponding to the first image;

[0030] A pixel point in the first image whose gray value is greater than or equal to the gray threshold corresponding to the first image is taken as a first pixel point; a pixel point in the first image whose gray value is less than the gray threshold corresponding to the first image is taken as a second pixel point;

[0031] A first computing unit is configured to:

[0032] Calculating a contrast evaluation value of the first image;

[0033] comparing the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, using the first image as the second image;

[0034] A second calculation unit is used to screen the pixel points in the second image to obtain a number of pixel points to be denoised corresponding to the second image;

[0035] The first noise reduction unit is used to reduce the noise of the plurality of pixels to be reduced in noise based on low-pass filtering to obtain a noise-reduced image to be processed; and traverse all the visual images of the patient to obtain a noise-reduced visual image.

[0036] Preferably, the first computing unit comprises:

[0037] The first computing subunit is configured to:

[0038] Randomly select a pixel point in the first image as a first target pixel point;

[0039] Determine a first target area with the first target pixel point as the center and the first preset distance as the radius;

[0040] Respectively obtain the number of first pixel points and second pixel points in the first target area;

[0041] Determine, based on the number of first pixels and second pixels in the first target area, a contrast evaluation value of the first target pixel corresponding to the first target area;

[0042] Traversing all pixel points in the first image to obtain a contrast evaluation value corresponding to each first target pixel point;

[0043] Taking the average of the contrast evaluation values ​​corresponding to all the pixels in the first image as the contrast evaluation value of the first image;

[0044] The first determining subunit is configured to compare the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, use the first image as the second image.

[0045] Preferably, the second computing unit comprises:

[0046] The second computing subunit is configured to:

[0047] Obtaining the gray level of each pixel in the second image;

[0048] Randomly select a pixel point in the second image as the second target pixel point;

[0049] Determine a second target area with the second target pixel point as the center and the second preset distance as the radius;

[0050] Determine the noise evaluation value corresponding to the second target pixel point based on the first preset algorithm according to each pixel point in the second target area;

[0051] Comparing the noise evaluation value corresponding to the second target pixel with a preset pixel noise evaluation threshold, and when it is determined that the noise evaluation value corresponding to the second target pixel is greater than or equal to the preset pixel noise evaluation threshold, using the second target pixel as a pixel to be denoised;

[0052] The second determining subunit is used to traverse all pixel points in the second image to obtain a number of pixel points to be denoised.

[0053] Preferably, the second noise reduction submodule includes:

[0054] Framing unit, used for:

[0055] Acquiring a speech signal from the patient's auditory data;

[0056] Performing frame processing on the speech signal to obtain a plurality of frames of speech sub-signals;

[0057] The third computing unit is configured to:

[0058] Calculate the noise evaluation value corresponding to each frame of speech sub-signal based on the second preset algorithm;

[0059] Secondary taxon for:

[0060] The noise evaluation value is compared with a preset noise threshold value, and if the noise evaluation value is greater than or equal to the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the speech frame; if the noise evaluation value is less than the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the noise frame;

[0061] Traverse each frame of speech sub-signal to obtain a noise frame set and a speech frame set;

[0062] A transform unit, used to perform Fourier transform on the speech sub-signals in the noise frame set and the speech frame set respectively, to obtain the noise frame frequency domain speech sub-signals corresponding to the noise frame set and the speech frame frequency domain speech sub-signals corresponding to the speech frame set;

[0063] Identify units for:

[0064] Acquire the spectrum energy corresponding to each noise frame frequency domain speech sub-signal to obtain the first spectrum energy;

[0065] Acquire the spectrum energy corresponding to the frequency-domain speech sub-signal of each speech frame to obtain the second spectrum energy;

[0066] The noise frame corresponding to the maximum value of the first spectrum energy is used as the first target frame;

[0067] Taking the speech frame corresponding to the maximum value of the second spectrum energy as the second target frame;

[0068] Determine the noise energy of the frequency domain speech signal corresponding to the speech signal based on the first target frame and the second target frame;

[0069] A query unit, configured to query a preset noise energy-filter coefficient table based on the noise energy to determine a filter coefficient corresponding to the speech signal;

[0070] The second noise reduction unit is used to reduce noise on the speech signal based on the filter coefficient to obtain noise-reduced audio data.

[0071] Preferably, the second preset algorithm includes:

[0072]

[0073] Among them, P i represents the noise evaluation value corresponding to the i-th frame speech sub-signal; x i，j represents the decibel value of the jth time domain sampling point in the i-th frame speech sub-signal; n represents the total number of time domain sampling points in the i-th frame speech sub-signal.

[0074] Preferably, the enhancer module comprises:

[0075] An acquisition unit, used for acquiring the cut data of the patient;

[0076] The enhancement unit is used to input the patient's cut data into a preset trained cut data enhancement model for enhancement to obtain enhanced cut data.

[0077] Preferably, the method for constructing an intelligent diagnosis model includes:

[0078] Obtain a training data set of multiple diagnostic methods including observation, auscultation, inquiry and palpation of patients;

[0079] The neural network model is trained based on the patient's multi-diagnosis training data set of observation, auscultation, questioning and palpation to obtain an initial intelligent diagnosis model;

[0080] Obtain a test data set of patients' multi-diagnosis including observation, auscultation, questioning and palpation;

[0081] The initial intelligent diagnosis model is tested based on a training data set of the patient's multi-diagnosis methods of observation, auscultation, questioning and palpation. When it is determined that the test results meet the requirements, a trained intelligent diagnosis model is obtained.

[0082] The present invention discloses a data processing system for TCM multi-diagnosis combining observation, auscultation, inquiry and palpation. Through automated data collection and processing, time and manpower are saved and diagnosis results can be obtained quickly. The multi-faceted data of observation, auscultation, inquiry and palpation are integrated and analyzed to make the diagnosis more comprehensive and accurate and avoid the one-sidedness that may be caused by a single method. The system is helpful to realize the standardized process of TCM diagnosis and reduce the interference and error of human factors. The system can continuously accumulate patient data and provide rich materials for subsequent medical research and experience summary, which is beneficial to the development and inheritance of TCM. The system provides intelligent diagnosis results for doctors as a reference and helps them to better formulate treatment plans.

[0083] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0084] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0086] Figure 1 is a block diagram of a data processing system for TCM multi-diagnosis based on observation, listening, questioning and palpation according to an embodiment of the present invention;

[0087] Figure 2 is a block diagram of a collection module according to an embodiment of the present invention;

[0088] Figure 3 is a block diagram of a preprocessing module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0089] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0090] Example 1

[0091] like Figure 1 As shown, a data processing system for TCM multi-diagnosis combined with observation, listening, questioning and palpation, comprises:

[0092] The acquisition module is used to collect the patient's multi-diagnosis data through observation, auscultation, questioning and palpation;

[0093] A preprocessing module is used to preprocess the multi-diagnosis combined reference data of the patient's inspection, auscultation, questioning and palpation to obtain target multi-diagnosis combined reference data;

[0094] An intelligent diagnosis module is used to input the target multi-diagnosis combined data into a pre-trained intelligent diagnosis model for analysis to obtain an intelligent diagnosis result;

[0095] The result output module is used to output the intelligent diagnosis results.

[0096] In this embodiment, the visual data includes but is not limited to the patient's facial expression, shape, tongue, skin and other images. The smell data includes but is not limited to the patient's voice, breathing, and cough data; the question data includes but is not limited to the patient's symptoms, medical history, living habits and related information; and the pulse data includes the patient's pulse data.

[0097] The beneficial effects of the above technical solution are: through automated data collection and processing, time and manpower are saved, and diagnostic results can be quickly obtained; data from multiple aspects of observation, auscultation, questioning and palpation are integrated and analyzed to make the diagnosis more comprehensive and accurate, avoiding the one-sidedness that may be caused by a single method; it helps to achieve a standardized process for Chinese medicine diagnosis and reduce interference and errors caused by human factors; it can continuously accumulate patient data, provide rich materials for subsequent medical research and experience summarization, and is conducive to the development and inheritance of Chinese medicine; it provides doctors with intelligent diagnostic results as a reference to help them better formulate treatment plans.

[0098] Example 2

[0099] like Figure 2 As shown, the acquisition module includes:

[0100] The first acquisition submodule is used to acquire images of the patient based on an image acquisition device to obtain a visual image of the patient;

[0101] The second acquisition submodule is used to collect the patient's voice information based on the voice acquisition device to obtain the patient's voice data;

[0102] The third collection submodule is used to obtain patient data based on a questionnaire;

[0103] The fourth acquisition submodule is used to collect the pulse data of the patient based on the sensor to obtain the patient's pulse data;

[0104] The first determination submodule is used to use the patient's visual image, the patient's smell data, the patient's question data and the patient's palpation data as the patient's visual, smell, question and palpation multi-diagnosis combined reference data.

[0105] In this embodiment, the image acquisition device includes but is not limited to a digital camera and a high-definition camera.

[0106] In this embodiment, the voice collection device includes but is not limited to a microphone, a professional recording device, etc.

[0107] In this embodiment, the questionnaire includes electronic and paper questionnaires.

[0108] The beneficial effects of the above technical solution are: it realizes multi-dimensional and comprehensive patient information collection, covering images, voice, questionnaires and physiological data, making the diagnosis and evaluation of patients more comprehensive and accurate; through different collection sub-modules, richer information can be obtained, which helps to have a deeper understanding of the patient's condition and physical condition; the use of modern technical means such as image acquisition equipment and voice acquisition equipment improves the efficiency and quality of information collection; and the integration of multiple data into multi-diagnosis and reference data provides a more solid foundation and a more comprehensive basis for subsequent medical diagnosis, treatment plan formulation, etc.

[0109] Example 3

[0110] like Figure 3 As shown, the preprocessing module includes:

[0111] A first denoising submodule, configured to denoise the patient's visual image to obtain a denoised visual image;

[0112] A second denoising submodule, used to denoise the patient's heard data to obtain denoised heard data;

[0113] A screening submodule, used for screening the patient's question data to obtain screened question data;

[0114] An enhancer module, used for enhancing the cut data of the patient to obtain enhanced hear data;

[0115] The second determination submodule is used to use the denoised visual image, denoised hearing data, screened patient question data and enhanced hearing data as target multi-diagnosis combined reference data.

[0116] In this embodiment, the screening submodule is used to screen the patient's question data to obtain screened question data; wherein the purpose of the screening is to remove invalid or redundant information.

[0117] The beneficial effects of the above technical solution are: through noise reduction processing, the quality and clarity of visual images and aural data are improved, noise interference is reduced, and subsequent analysis is more accurate; the question data is screened to remove invalid or redundant information, making the data more targeted and effective; the cross-data is enhanced to highlight key information, which helps to more accurately analyze the patient's physiological condition; the various types of processed data are integrated to form target multi-diagnosis data, providing a more reliable and valuable data basis for further diagnosis and treatment, and improving the scientificity and accuracy of medical decision-making; the overall availability and reliability of data are improved, which is conducive to improving the quality and level of medical services.

[0118] Example 4

[0119] The first denoising submodule includes:

[0120] First classification unit for:

[0121] Take any patient's visual image as the image to be processed;

[0122] Grayscale the image to be processed to obtain a grayscale image;

[0123] The grayscale image is evenly divided into a number of sub-images to obtain a number of grayscale sub-images;

[0124] Take any grayscale sub-image as the first image;

[0125] Obtaining the grayscale value of each pixel in the first image;

[0126] Based on the Otsu threshold segmentation method, determine the grayscale threshold corresponding to the first image;

[0127] Compare the grayscale value of each pixel in the first image with the grayscale threshold corresponding to the first image;

[0128] A pixel point in the first image whose gray value is greater than or equal to the gray threshold corresponding to the first image is taken as a first pixel point; a pixel point in the first image whose gray value is less than the gray threshold corresponding to the first image is taken as a second pixel point;

[0129] A first computing unit is configured to:

[0130] Calculating a contrast evaluation value of the first image;

[0131] comparing the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, using the first image as the second image;

[0132] A second calculation unit is used to screen the pixel points in the second image to obtain a number of pixel points to be denoised corresponding to the second image;

[0133] The first noise reduction unit is used to reduce the noise of the plurality of pixels to be reduced in noise based on low-pass filtering to obtain a noise-reduced image to be processed; and traverse all the visual images of the patient to obtain a noise-reduced visual image.

[0134] In this embodiment, Otsu threshold segmentation, also called maximum inter-class variance method, is a method for automatically determining the image segmentation threshold.

[0135] In this embodiment, the preset contrast evaluation threshold is set in advance based on industry experience.

[0136] In this embodiment, the first image whose contrast evaluation value is greater than or equal to the preset contrast evaluation threshold is selected as the second image, indicating that the contrast of the first image is too large, resulting in image distortion or partial detail loss, and therefore it is selected as the image to be denoised.

[0137] The working principle of the above technical solution is: a first classification unit is used to select a patient's visual image and grayscale it, and then it is divided into multiple sub-images, and then a sub-image is selected as the first image, and its pixel grayscale value is obtained. Then, the grayscale threshold of the first image is determined by Otsu threshold segmentation, so as to distinguish the first pixel and the second pixel; the first calculation unit calculates the contrast evaluation value of the first image and compares it with the preset threshold, and the first image that meets the conditions is determined as the second image; the second calculation unit screens the pixels in the second image to find the pixels that need to be denoised; the first noise reduction unit uses low-pass filtering technology to process these pixels to be denoised to achieve a noise reduction effect, and by repeating this process for all visual images, the overall visual image after noise reduction is finally obtained.

[0138] The beneficial effects of the above technical solution are: through a series of operations such as grayscale, segmentation, threshold processing, and noise reduction, the noise in the image can be effectively reduced, making the image clearer and facilitating subsequent analysis and diagnosis; screening out the pixels to be denoised and performing targeted processing is helpful to retain important details of the image and improve the accuracy and readability of the image; uniformly segmenting the image and processing it step by step, and through specific calculation and screening methods, while ensuring the effect, the overall processing efficiency is improved; based on Otsu threshold segmentation and preset thresholds, etc., it can adapt to the image characteristics of different patients and has certain versatility and flexibility; high-quality denoised images are obtained, providing a more reliable data basis for further medical diagnosis, disease assessment, etc.

[0139] Example 5

[0140] The first computing unit comprises:

[0141] The first computing subunit is configured to:

[0142] Randomly select a pixel point in the first image as a first target pixel point;

[0143] Determine a first target area with the first target pixel point as the center and the first preset distance as the radius;

[0144] Respectively obtain the number of first pixel points and second pixel points in the first target area;

[0145] Determine, based on the number of first pixels and second pixels in the first target area, a contrast evaluation value of the first target pixel corresponding to the first target area;

[0146] Traversing all pixel points in the first image to obtain a contrast evaluation value corresponding to each first target pixel point;

[0147] Taking the average of the contrast evaluation values ​​corresponding to all the pixels in the first image as the contrast evaluation value of the first image;

[0148] The first determining subunit is configured to compare the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, use the first image as the second image.

[0149] In this embodiment, based on the number of first pixels and second pixels in the first target area, a specific implementation method for determining the contrast evaluation value of the first target pixel corresponding to the first target area is as follows: the ratio of the number of first pixels to the number of second pixels in the first target area is used as the contrast evaluation value of the first target pixel corresponding to the first target area.

[0150] The beneficial effects of the above technical solution are: by determining the target area with each pixel as the center and analyzing the number of different pixels therein to calculate the contrast evaluation value, the local contrast situation can be reflected more accurately; all pixels in the image are traversed, so that the obtained contrast evaluation value is more representative and comprehensive; based on the comparison between the contrast evaluation value and the preset threshold, images that meet the specific contrast requirements can be accurately screened out, providing more suitable objects for subsequent processing; this calculation method can adapt to different image features and scenes, improving the adaptability and flexibility of the solution; the determined contrast evaluation value helps to classify, screen and further process the image more reasonably, and improve the overall processing effect and quality.

[0151] Example 6

[0152] The second computing unit comprises:

[0153] The second computing subunit is configured to:

[0154] Obtaining the gray level of each pixel in the second image;

[0155] Randomly select a pixel point in the second image as the second target pixel point;

[0156] Determine a second target area with the second target pixel point as the center and the second preset distance as the radius;

[0157] Determine the noise evaluation value corresponding to the second target pixel point based on the first preset algorithm according to each pixel point in the second target area;

[0158] Comparing the noise evaluation value corresponding to the second target pixel with a preset pixel noise evaluation threshold, and when it is determined that the noise evaluation value corresponding to the second target pixel is greater than or equal to the preset pixel noise evaluation threshold, using the second target pixel as a pixel to be denoised;

[0159] The second determining subunit is used to traverse all pixel points in the second image to obtain a number of pixel points to be denoised.

[0160] In this embodiment, the first preset algorithm includes:

[0161]

[0162] Among them, V t represents the noise evaluation value corresponding to the t-th target pixel; H(s t ) represents the information entropy of all grayscale differences in the second target area corresponding to the t-th target pixel; n represents the total number of grayscale levels in the second target area corresponding to the t-th target pixel; s r，t represents the rth gray level in the second target area corresponding to the tth target pixel; P(s r，t) represents the frequency of occurrence of the rth gray level in the second target area corresponding to the tth target pixel.

[0163] In this embodiment, the preset pixel noise evaluation threshold is set in advance based on industry experience.

[0164] The beneficial effects of the above technical solution are: by calculating the noise evaluation value through the setting of a specific algorithm and the target area, the pixels that may have noise can be located more accurately; the pixels to be denoised are screened out, and targeted noise reduction processing is achieved, avoiding excessive processing of all pixels, which helps to retain the effective information of the image; the quality of the image after noise reduction can be better improved, making the image clearer; the preset threshold and other parameters can be adjusted according to the characteristics of different images, and it has good adaptability; while reducing noise, the loss of image details can be minimized to maintain the integrity and availability of the image.

[0165] Example 7

[0166] The second denoising submodule includes:

[0167] Framing unit, used for:

[0168] Acquiring a speech signal from the patient's auditory data;

[0169] Performing frame processing on the speech signal to obtain a plurality of frames of speech sub-signals;

[0170] The third computing unit is configured to:

[0171] Calculate the noise evaluation value corresponding to each frame of speech sub-signal based on the second preset algorithm;

[0172] Secondary taxon for:

[0173] The noise evaluation value is compared with a preset noise threshold value, and if the noise evaluation value is greater than or equal to the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the speech frame; if the noise evaluation value is less than the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the noise frame;

[0174] Traverse each frame of speech sub-signal to obtain a noise frame set and a speech frame set;

[0175] A transform unit, used to perform Fourier transform on the speech sub-signals in the noise frame set and the speech frame set respectively, to obtain the noise frame frequency domain speech sub-signals corresponding to the noise frame set and the speech frame frequency domain speech sub-signals corresponding to the speech frame set;

[0176] Identify units for:

[0177] Acquire the spectrum energy corresponding to each noise frame frequency domain speech sub-signal to obtain the first spectrum energy;

[0178] Acquire the spectrum energy corresponding to the frequency-domain speech sub-signal of each speech frame to obtain the second spectrum energy;

[0179] The noise frame corresponding to the maximum value of the first spectrum energy is used as the first target frame;

[0180] Taking the speech frame corresponding to the maximum value of the second spectrum energy as the second target frame;

[0181] Determine the noise energy of the frequency domain speech signal corresponding to the speech signal based on the first target frame and the second target frame;

[0182] A query unit, configured to query a preset noise energy-filter coefficient table based on the noise energy to determine a filter coefficient corresponding to the speech signal;

[0183] The second noise reduction unit is used to reduce noise on the speech signal based on the filter coefficient to obtain noise-reduced audio data.

[0184] In this embodiment, the preset noise threshold is set in advance based on industry experience.

[0185] In this embodiment, the preset noise energy-filter coefficient table is set in advance based on industry experience of historical data sets.

[0186] The beneficial effects of the above technical solution are: by dividing the frame and calculating the noise evaluation value, the speech frame and the noise frame can be accurately distinguished, providing a basis for subsequent targeted processing; by determining the noise energy based on the spectral energy and obtaining the appropriate filter coefficient, more effective noise reduction processing can be achieved and the quality of the speech signal can be improved; in the noise reduction process, through reasonable processing, the important details and characteristics of the speech signal are retained as much as possible; it can be flexibly adjusted and processed according to the speech data of different patients, and has good adaptability; through the detailed analysis of the noise frame and the speech frame, the accuracy and reliability of the speech signal processing are improved.

[0187] Example 8

[0188] The second preset algorithm includes:

[0189]

[0190] Among them, P i represents the noise evaluation value corresponding to the i-th frame speech sub-signal; x i，j represents the decibel value of the jth time domain sampling point in the i-th frame speech sub-signal; n represents the total number of time domain sampling points in the i-th frame speech sub-signal.

[0191] The beneficial effects of the above technical solution are: it fully considers the information of each time domain sampling point in each frame of speech sub-signal, and can better reflect the characteristics of the signal; it can more accurately quantify the noise level of each frame of speech sub-signal, providing a basis for subsequent classification and processing.

[0192] Example 9

[0193] Enhancer modules, including:

[0194] An acquisition unit, used for acquiring the cut data of the patient;

[0195] The enhancement unit is used to input the patient's cut data into a preset trained cut data enhancement model for enhancement to obtain enhanced cut data.

[0196] In this embodiment, the method for constructing the data enhancement model includes:

[0197] Get the data set for data cutting training;

[0198] Training the neural network model based on the cut data training data set to obtain an initial cut data enhancement model;

[0199] Get the data set for testing;

[0200] The initial data-cutting enhancement model is tested based on the data-cutting test data set, and when it is determined that the test result meets the requirements, a trained data-cutting enhancement model is obtained.

[0201] The beneficial effects of the above technical solution are: it can make up for the deficiencies or defects that may exist in the original cut data and improve the integrity and accuracy of the data; by enhancing the operation, more diverse cut data can be generated, which helps to more comprehensively reflect various possible situations, thereby improving the reliability of subsequent analysis and processing; and enable applications based on the cut data to better cope with a variety of actual scenarios and situation changes.

[0202] Example 10

[0203] The method for constructing an intelligent diagnosis model includes:

[0204] Obtain a training data set of multiple diagnostic methods including observation, auscultation, inquiry and palpation of patients;

[0205] The neural network model is trained based on the patient's multi-diagnosis training data set of observation, auscultation, questioning and palpation to obtain an initial intelligent diagnosis model;

[0206] Obtain a test data set of patients' multi-diagnosis including observation, auscultation, questioning and palpation;

[0207] The initial intelligent diagnosis model is tested based on a training data set of the patient's multi-diagnosis methods of observation, auscultation, questioning and palpation. When it is determined that the test results meet the requirements, a trained intelligent diagnosis model is obtained.

[0208] The beneficial effects of the above technical solution are: combining data from multiple aspects such as observation, auscultation, questioning and palpation can analyze and diagnose the patient's condition more comprehensively and systematically; using neural network models for training and optimization can help improve the accuracy and reliability of diagnosis; it can analyze the specific multi-diagnosis data of different patients to provide more personalized diagnosis results; based on a large number of training data sets, diagnostic decisions are more based on evidence and reduce subjective bias; and it provides doctors with powerful auxiliary tools to improve medical efficiency and quality.

[0209] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A data processing system for TCM multi-diagnosis combined with observation, auscultation, inquiry and palpation, characterized in that: include: The acquisition module is used to collect the patient's multi-diagnosis data through observation, auscultation, questioning and palpation; A preprocessing module is used to preprocess the multi-diagnosis combined reference data of the patient's inspection, auscultation, questioning and palpation to obtain target multi-diagnosis combined reference data; An intelligent diagnosis module is used to input the target multi-diagnosis combined data into a pre-trained intelligent diagnosis model for analysis to obtain an intelligent diagnosis result; The result output module is used to output the intelligent diagnosis results.

2. The TCM multi-diagnosis data processing system according to claim 1, characterized in that: The acquisition module comprises: The first acquisition submodule is used to acquire images of the patient based on an image acquisition device to obtain a visual image of the patient; The second acquisition submodule is used to collect the patient's voice information based on the voice acquisition device to obtain the patient's voice data; The third collection submodule is used to obtain patient data based on a questionnaire; The fourth acquisition submodule is used to collect the pulse data of the patient based on the sensor to obtain the patient's pulse data; The first determination submodule is used to use the patient's visual image, the patient's smell data, the patient's question data and the patient's palpation data as the patient's visual, smell, question and palpation multi-diagnosis combined reference data.

3. The TCM multi-diagnosis data processing system according to claim 2, characterized in that: The preprocessing module comprises: A first denoising submodule, configured to denoise the patient's visual image to obtain a denoised visual image; A second denoising submodule, used to denoise the patient's heard data to obtain denoised heard data; A screening submodule, used for screening the patient's question data to obtain screened question data; An enhancer module, used for enhancing the cut data of the patient to obtain enhanced hear data; The second determination submodule is used to use the denoised visual image, denoised hearing data, screened patient question data and enhanced hearing data as target multi-diagnosis combined reference data.

4. The TCM multi-diagnosis data processing system according to claim 3, characterized in that: The first denoising submodule includes: First classification unit for: Take any patient's visual image as the image to be processed; Grayscale the image to be processed to obtain a grayscale image; The grayscale image is evenly divided into a number of sub-images to obtain a number of grayscale sub-images; Take any grayscale sub-image as the first image; Obtaining the grayscale value of each pixel in the first image; Based on the Otsu threshold segmentation method, determine the grayscale threshold corresponding to the first image; Compare the grayscale value of each pixel in the first image with the grayscale threshold corresponding to the first image; A pixel point in the first image whose gray value is greater than or equal to the gray threshold corresponding to the first image is taken as a first pixel point; a pixel point in the first image whose gray value is less than the gray threshold corresponding to the first image is taken as a second pixel point; A first computing unit is configured to: Calculating a contrast evaluation value of the first image; comparing the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, using the first image as the second image; A second calculation unit is used to screen the pixel points in the second image to obtain a number of pixel points to be denoised corresponding to the second image; The first noise reduction unit is used to reduce the noise of the plurality of pixels to be reduced in noise based on low-pass filtering to obtain a noise-reduced image to be processed; and traverse all the visual images of the patient to obtain a noise-reduced visual image.

5. The TCM multi-diagnosis data processing system according to claim 4, characterized in that: The first computing unit comprises: The first computing subunit is configured to: Randomly select a pixel point in the first image as a first target pixel point; Determine a first target area with the first target pixel point as the center and the first preset distance as the radius; Respectively obtain the number of first pixel points and second pixel points in the first target area; Determine, based on the number of first pixels and second pixels in the first target area, a contrast evaluation value of the first target pixel corresponding to the first target area; Traversing all pixel points in the first image to obtain a contrast evaluation value corresponding to each first target pixel point; Taking the average of the contrast evaluation values ​​corresponding to all the pixels in the first image as the contrast evaluation value of the first image; The first determining subunit is configured to compare the contrast evaluation value of the first image with a preset contrast evaluation threshold, and when it is determined that the contrast evaluation value of the first image is greater than or equal to the preset contrast evaluation threshold, use the first image as the second image.

6. The TCM multi-diagnosis data processing system according to claim 4, characterized in that: The second computing unit comprises: The second computing subunit is configured to: Obtaining the gray level of each pixel in the second image; Randomly select a pixel point in the second image as the second target pixel point; Determine a second target area with the second target pixel point as the center and the second preset distance as the radius; Determine the noise evaluation value corresponding to the second target pixel point based on the first preset algorithm according to each pixel point in the second target area; Comparing the noise evaluation value corresponding to the second target pixel with a preset pixel noise evaluation threshold, and when it is determined that the noise evaluation value corresponding to the second target pixel is greater than or equal to the preset pixel noise evaluation threshold, using the second target pixel as a pixel to be denoised; The second determining subunit is used to traverse all pixel points in the second image to obtain a number of pixel points to be denoised.

7. The TCM multi-diagnosis data processing system according to claim 3, characterized in that: The second denoising submodule includes: Framing unit, used for: Acquiring a speech signal from the patient's auditory data; Performing frame processing on the speech signal to obtain a plurality of frames of speech sub-signals; The third computing unit is configured to: Calculate the noise evaluation value corresponding to each frame of speech sub-signal based on the second preset algorithm; Secondary taxon for: The noise evaluation value is compared with a preset noise threshold value, and if the noise evaluation value is greater than or equal to the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the speech frame; if the noise evaluation value is less than the preset noise threshold value, the speech sub-signal corresponding to the noise evaluation value is used as the speech sub-signal of the noise frame; Traverse each frame of speech sub-signal to obtain a noise frame set and a speech frame set; A transform unit, used to perform Fourier transform on the speech sub-signals in the noise frame set and the speech frame set respectively, to obtain the noise frame frequency domain speech sub-signals corresponding to the noise frame set and the speech frame frequency domain speech sub-signals corresponding to the speech frame set; Identify units for: Acquire the spectrum energy corresponding to each noise frame frequency domain speech sub-signal to obtain the first spectrum energy; Acquire the spectrum energy corresponding to the frequency-domain speech sub-signal of each speech frame to obtain the second spectrum energy; The noise frame corresponding to the maximum value of the first spectrum energy is used as the first target frame; Taking the speech frame corresponding to the maximum value of the second spectrum energy as the second target frame; Determine the noise energy of the frequency domain speech signal corresponding to the speech signal based on the first target frame and the second target frame; A query unit, configured to query a preset noise energy-filter coefficient table based on the noise energy to determine a filter coefficient corresponding to the speech signal; The second noise reduction unit is used to reduce noise on the speech signal based on the filter coefficient to obtain noise-reduced audio data.

8. The TCM multi-diagnosis data processing system according to claim 7, characterized in that: The second preset algorithm includes: Among them, P i represents the noise evaluation value corresponding to the i-th frame speech sub-signal; x i，j represents the decibel value of the jth time domain sampling point in the i-th frame speech sub-signal; n represents the total number of time domain sampling points in the i-th frame speech sub-signal.

9. The TCM multi-diagnosis data processing system according to claim 3, characterized in that: Enhancer modules, including: An acquisition unit, used for acquiring the cut data of the patient; The enhancement unit is used to input the patient's cut data into a preset trained cut data enhancement model for enhancement to obtain enhanced cut data.

10. The TCM multi-diagnosis data processing system according to claim 1, characterized in that: The method for constructing an intelligent diagnosis model includes: Obtain a training data set of multiple diagnostic methods including observation, auscultation, inquiry and palpation of patients; The neural network model is trained based on the patient's multi-diagnosis training data set of observation, auscultation, questioning and palpation to obtain an initial intelligent diagnosis model; Obtain a test data set of patients' multi-diagnosis including observation, auscultation, questioning and palpation; The initial intelligent diagnosis model is tested based on a training data set of the patient's multi-diagnosis methods of observation, auscultation, questioning and palpation. When it is determined that the test results meet the requirements, a trained intelligent diagnosis model is obtained.

Images