Parkinson's disease intelligent diagnosis system based on finger pressure identification

Abstract

The invention discloses a Parkinson's disease intelligent diagnosis system based on finger pressure recognition, and aims to solve the problems that in the prior art, behavior characteristics related to Parkinson's disease are difficult to comprehensively capture only by depending on a finger pressure change mode, individual differences of patients are not fully considered, and the diagnosis accuracy is poor. The diagnosis result is easily influenced by data noise or individual characteristics, so that the accuracy is influenced; through comprehensive acquisition and analysis of multi-dimensional data, the early diagnosis capability of the Parkinson's disease is remarkably improved, behavior characteristics related to the Parkinson's disease can be comprehensively captured, symptoms such as tremor, motion retardation and force fluctuation can be accurately recognized, an accurate behavior characteristic report is provided, and the diagnosis accuracy of the Parkinson's disease is improved by combining historical clinical data and a set threshold value. The comprehensive abnormal score can prompt possible symptoms of the Parkinson's disease in time, assists a doctor in making a decision, ensures the accuracy and individuation of diagnosis, and provides important support for early intervention and precise treatment of the Parkinson's disease.

Description

Technical Field

[0001] The invention relates to the technical field of Parkinson's disease diagnosis, and in particular to an intelligent Parkinson's disease diagnosis system based on finger pressure recognition. Background Art

[0002] The early symptoms of Parkinson's disease patients are relatively mild, and the appearance of various symptoms does not show regularity. In addition, the current diagnosis mostly relies on the doctor's observation and evaluation, resulting in a low diagnosis rate and a high misdiagnosis rate. The patent application with publication number: CN113729715A discloses an intelligent diagnosis system for Parkinson's disease based on finger pressure data, including: a data acquisition device for obtaining the pressure data of five fingers of the tester for a period of time when the tester grasps and sends it to a single-chip microcomputer; a single-chip microcomputer for obtaining the probability of illness based on the five finger pressure data according to the pre-trained diagnostic model, and fusing the obtained five probabilities of illness to obtain a diagnostic conclusion, wherein the diagnostic model is obtained based on the training of a convolutional neural network. By collecting the pressure data of the five fingers of the diagnosed person and combining the deep convolutional neural network to diagnose Parkinson's disease, the system has a high degree of intelligence and saves costs.

[0003] Although the above patent can make a preliminary judgment based on finger pressure data, it only relies on the pressure change pattern of the fingers, which makes it difficult to fully capture the behavioral characteristics related to Parkinson's disease, and does not fully consider the individual differences of patients, resulting in the diagnosis results being easily affected by data noise or individual characteristics, thereby affecting accuracy. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent Parkinson's disease diagnosis system based on finger pressure recognition, which improves the accuracy of diagnosis, reduces the misdiagnosis rate, and provides patients with more accurate health assessments and personalized treatment recommendations through multi-dimensional data collection and comprehensive analysis. It also optimizes the diagnosis process, improves the quality and efficiency of medical services, and provides important support for early intervention and precise treatment of Parkinson's disease through precise monitoring, personalized treatment recommendations, and data security assurance, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: The Parkinson's disease intelligent diagnosis system based on finger pressure recognition includes: A biometric feature collection unit is used to collect the patient's finger pressure data based on fingerprint recognition and finger pressure sensors, and at the same time, obtain the patient's biometric data and verify and screen the patient's identity; A pressure data analysis unit, used to monitor dynamic pressure changes based on the patient's finger pressure data, and to determine whether the patient has Parkinson's disease-related behavioral characteristics by analyzing the dynamic pressure changes of the fingers; The Parkinson's disease diagnosis unit is used to integrate the collected biometric data, behavioral characteristics and dynamic pressure analysis data, analyze the possible symptoms of Parkinson's disease, compare them with the medical expertise database, generate diagnosis results, and provide auxiliary diagnosis suggestions.

[0006] Furthermore, the biometric feature collection unit includes: The fingerprint recognition module is used to obtain the patient's fingerprint image data in real time based on the fingerprint sensor, and pre-process the collected fingerprint image data to extract the patient's fingerprint features from the processed fingerprint image; The pressure sensing module is used to collect the pressure distribution on the patient's fingers through the finger pressure sensor. At the same time, each finger is labeled based on the patient's fingerprint characteristics, and a corresponding pressure change data set is generated based on each finger label.

[0007] Furthermore, the fingerprint recognition module includes: An image extraction module, used to extract the patient's fingerprint image data; A region division module, used for obtaining a fingerprint region and a non-fingerprint region from the fingerprint image data; A first grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixels contained in the fingerprint area as a first grayscale value set; A second grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixel points included in the non-fingerprint area as a second grayscale value set; A first gray value coefficient acquisition module, used to acquire a first gray value coefficient by using the gray value of the pixel point included in the first gray value set; The first gray value coefficient is obtained by the following formula: Among them, J 01 represents the first gray value coefficient; n represents the number of pixels contained in the first gray value set; H 01i represents the gray value of the i-th pixel point included in the first gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H c Indicates the preset grayscale reference value; H 01bi Represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the first grayscale value set; A second gray value coefficient acquisition module, used to acquire a second gray value coefficient by using the gray value of the pixel point included in the second gray value set; The second gray value coefficient is obtained by the following formula: Among them, J 02 represents the second gray value coefficient; m represents the number of pixels included in the second gray value set; H 02i represents the gray value of the i-th pixel point included in the second gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H 02bi represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the second grayscale value set; A coefficient difference acquisition module, used for comparing the first gray value coefficient with the second gray value coefficient to acquire the coefficient difference between the first gray value coefficient and the second gray value coefficient; A coefficient difference comparison module, used for comparing the coefficient difference with a preset coefficient difference threshold; The contrast adjustment determination module is used to perform contrast adjustment on the fingerprint image data when the coefficient difference is lower than a preset coefficient difference threshold.

[0008] Furthermore, the contrast adjustment determination module includes: A coefficient difference retrieving module, used for retrieving the coefficient difference between the first gray value coefficient and the second gray value coefficient when the coefficient difference is lower than a preset coefficient difference threshold; A first contrast adjustment module, configured to adjust the contrast of the fingerprint region by using a coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the first gray value coefficient; The adjusted contrast value corresponding to the fingerprint area is obtained by the following formula: Among them, D z Indicates the adjusted contrast value corresponding to the fingerprint area; D z0 Indicates the contrast value before adjustment corresponding to the fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold; A second contrast adjustment module, configured to adjust the contrast of the non-fingerprint area by using the coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the second gray value coefficient; The adjusted contrast value corresponding to the non-fingerprint area is obtained by the following formula: Among them, D fIndicates the adjusted contrast value corresponding to the non-fingerprint area; D f0 Indicates the contrast value before adjustment corresponding to the non-fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold.

[0009] Furthermore, the biometric feature collection unit further includes: The identity verification module is used to compare the extracted fingerprint features with the patient identity data stored in the patient database to verify the patient's identity information. If the verification fails, an identity abnormality warning is issued and the monitoring operation is terminated; The file management module is used to associate the extracted fingerprint feature data with the patient's identity information, construct the patient's personal file data, and manage and store the acquired patient's personal file data; The data storage module is used to store the collected patient biometric data, health records and diagnostic data, update the patient's personal file data after each visit, and synchronize the updated personal file data to the database in real time.

[0010] Furthermore, the pressure data analysis unit comprises: The pressure change monitoring module is used to monitor the dynamic changes of the patient's finger pressure data based on the time series of the pressure change data set, perform fluctuation analysis on the pressure data in each time period, and generate a pressure change report; The behavior feature analysis module is used to extract the behavior features associated with finger movement from the finger pressure change data, and combine the time series of pressure changes to determine the reaction speed of the patient's fingers during different movements and evaluate the degree of bradykinesia; The arteriovenous recognition module is used to capture the vein map of the patient's finger based on infrared imaging equipment, and distinguish between live and non-live samples in the patient's finger pressure data in combination with liveness detection technology; The behavior feature analysis module combines the recognition result of the arteriovenous recognition module and issues a false data abnormality alarm when it recognizes that there are non-living samples in the finger pressure data.

[0011] Furthermore, the behavior feature analysis module specifically includes: Perform frequency domain analysis on finger pressure data to extract tremor frequency and amplitude, determine the frequency peak based on the extraction results, identify whether the patient has tremor behavior, calculate the tremor amplitude, and determine the patient's tremor frequency; Based on the time series of pressure changes, the time delay of each pressure fluctuation cycle is calculated, the patient's reaction time is evaluated based on the degree of time delay, and the patient's bradykinesia index is determined; Analyze the fluctuation of pressure data during continuous force application, quantify the force fluctuation through standard deviation, maximum value and minimum value difference, and determine the patient's finger force fluctuation; The pressure data and behavioral characteristics of each action process are integrated in time series, the time axis is aligned, and the reaction speed and relative changes of the action are detected; Calculate the reaction time of pressure data for different finger movements to assess whether the patient has delayed behavior; Based on the time series, the changing trends of the behavioral characteristics of the fingers under different movements are quantified, and the abnormal scores of the behavioral characteristics are generated, and the patient's behavioral characteristics report is generated, including tremor amplitude, bradykinesia index, and force fluctuation.

[0012] Furthermore, the Parkinson's disease diagnosis unit includes: The comprehensive diagnosis module is used to compare the comprehensive abnormality score with the set threshold of the clinical diagnosis standard for Parkinson's disease. If the comprehensive score exceeds the threshold, it indicates that there may be symptoms related to Parkinson's disease, which assists doctors in making diagnostic decisions; The diagnosis result generation module is used to generate diagnostic suggestions based on the comprehensive abnormality score and patient historical data to assist doctors in determining whether the patient has symptoms of Parkinson's disease.

[0013] Furthermore, the comprehensive abnormality score specifically includes: The finger pressure data, behavioral characteristics, vein map and liveness detection results are weightedly fused to generate a unified multi-dimensional feature vector, and the time series characteristics of the multi-dimensional feature vector are extracted; Classify the fused multi-dimensional feature vectors to identify behavioral features associated with Parkinson's disease. At the same time, generate abnormal scores for each behavioral feature based on the classification results, including abnormal scores for tremor, bradykinesia, and force fluctuations. The abnormal scores are determined based on comparison with historical clinical data. By weighted summing up the abnormal scores of each behavioral feature, a comprehensive abnormal score for each patient is generated, and combined with the set threshold, it is determined whether there is a Parkinson's disease-related behavioral pattern; Based on the patient's historical diagnosis and treatment data, it is determined whether the current abnormal behavior characteristics are new symptoms or changes in existing symptoms, and targeted treatment recommendations are provided to doctors.

[0014] Furthermore, the Parkinson's disease diagnosis unit also includes: feeding back the patient's biometric data and diagnosis results to users and medical personnel through a human-computer interaction interface, generating a health report for the patient, and providing the patient with corresponding treatment suggestions, lifestyle adjustment plans and further medical examination instructions based on the diagnosis results and auxiliary diagnosis suggestions.

[0015] Compared with the prior art, the present invention has the following beneficial effects: By combining multiple biometric features such as fingerprint recognition, finger pressure data, and dynamic pressure analysis, the comprehensive collection and analysis of multi-dimensional data can effectively capture the behavioral characteristics related to Parkinson's disease, overcome the limitations of traditional methods, and comprehensively and accurately assess the patient's condition and reduce the misdiagnosis rate. At the same time, through personalized diagnosis and treatment recommendations and treatment plans, more accurate health management is provided to patients. The authenticity and security of the data are ensured, the privacy of patients is effectively protected, and the diagnosis process is optimized, improving the accuracy and efficiency of diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a module diagram of the Parkinson's disease intelligent diagnosis system based on finger pressure recognition of the present invention. DETAILED DESCRIPTION

[0017] 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.

[0018] In order to solve the technical problem that the existing technology is difficult to fully capture the behavioral characteristics related to Parkinson's disease, and does not fully consider the individual differences of patients, resulting in the diagnosis results being easily affected by data noise or individual characteristics, thus affecting the accuracy, please refer to Figure 1 , this embodiment provides the following technical solutions: The Parkinson's disease intelligent diagnosis system based on finger pressure recognition includes: The biometrics collection unit is used to collect the patient's finger pressure data based on fingerprint recognition and finger pressure sensors. At the same time, it obtains the patient's biometric data and verifies and screens the patient's identity to ensure the accuracy and authenticity of the data source; A pressure data analysis unit, used to monitor dynamic pressure changes based on the patient's finger pressure data, and to determine whether the patient has Parkinson's disease-related behavioral characteristics by analyzing the dynamic pressure changes of the fingers; The Parkinson's disease diagnosis unit is used to integrate the collected biometric data, behavioral characteristics and dynamic pressure analysis data, analyze the possible symptoms of Parkinson's disease, compare them with the medical professional knowledge database, generate diagnosis results, and provide auxiliary diagnosis suggestions; The patient's biometric data and diagnostic results are fed back to users and medical personnel through a human-computer interaction interface to generate a health report for the patient. Based on the diagnostic results and auxiliary diagnostic suggestions, the patient is provided with corresponding treatment recommendations, lifestyle adjustment plans and further medical examination guidance.

[0019] In this embodiment, by combining multiple biometric features such as fingerprint recognition, finger pressure data, dynamic pressure analysis, etc., the comprehensive collection and analysis of multi-dimensional data can effectively capture the behavioral characteristics related to Parkinson's disease, overcome the limitations of traditional methods, and comprehensively and accurately assess the patient's condition, reduce the misdiagnosis rate, and further ensure the authenticity and security of the data, effectively protect the privacy of the patient, and optimize the diagnosis process, improve the accuracy and efficiency of the diagnosis. At the same time, through personalized diagnosis and treatment recommendations and treatment plans, more accurate health management is provided for patients, the quality and efficiency of medical services are improved, and important support is provided for the early intervention and precision treatment of Parkinson's disease.

[0020] In this embodiment, the biometric feature collection unit includes: The fingerprint recognition module is used to obtain the patient's fingerprint image data in real time based on the fingerprint sensor, and pre-process the collected fingerprint image data to extract the patient's fingerprint features from the processed fingerprint image; The pressure sensing module is used to collect the pressure distribution on the patient's fingers through the finger pressure sensor, and at the same time, mark each finger based on the patient's fingerprint characteristics, and generate a corresponding pressure change data set based on each finger number mark; The identity verification module is used to compare the extracted fingerprint features with the patient identity data stored in the patient database to verify the patient's identity information. If the verification fails, an identity abnormality warning is issued and the monitoring operation is terminated; The file management module is used to associate the extracted fingerprint feature data with the patient's identity information, build the patient's personal file data, and manage and store the acquired patient's personal file data, including personal information, biometric data, health records and historical diagnosis results, to ensure the integrity and traceability of patient data, and support rapid retrieval and update of data to facilitate subsequent analysis and diagnosis by doctors or systems; The data storage module is used to store the collected patient biometric data, health records and diagnostic data, update the patient's personal file data after each visit, and synchronize the updated personal file data to the database in real time.

[0021] In this embodiment, each finger is labeled according to the patient's fingerprint features, and a corresponding pressure change data set is generated, which helps to more accurately monitor and analyze the pressure status of each finger, providing important information for diagnosis and treatment. It integrates multiple modules such as fingerprint recognition, pressure sensing, identity authentication, file management and data storage, and realizes accurate patient identity authentication, personalized finger pressure monitoring, real-time identity abnormality warning, comprehensive patient file management, and secure storage and synchronization of data, thereby improving the quality and efficiency of medical services, promoting the construction of medical informatization, enhancing patient experience, and ensuring the integrity and traceability of patient data, providing important support for building a smart medical system.

[0022] Specifically, the fingerprint recognition module includes: An image extraction module, used to extract the patient's fingerprint image data; A region division module, used for obtaining a fingerprint region and a non-fingerprint region from the fingerprint image data; A first grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixels contained in the fingerprint area as a first grayscale value set; A second grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixel points included in the non-fingerprint area as a second grayscale value set; A first gray value coefficient acquisition module, used to acquire a first gray value coefficient by using the gray value of the pixel point included in the first gray value set; The first gray value coefficient is obtained by the following formula: Among them, J 01 represents the first gray value coefficient; n represents the number of pixels contained in the first gray value set; H 01i represents the gray value of the i-th pixel point included in the first gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H c Indicates the preset grayscale reference value; H 01bi Represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the first grayscale value set; A second gray value coefficient acquisition module, used to acquire a second gray value coefficient by using the gray value of the pixel point included in the second gray value set; The second gray value coefficient is obtained by the following formula: Among them, J 02represents the second gray value coefficient; m represents the number of pixels included in the second gray value set; H 02i represents the gray value of the i-th pixel point included in the second gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H 02bi represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the second grayscale value set; A coefficient difference acquisition module, used for comparing the first gray value coefficient with the second gray value coefficient to acquire the coefficient difference between the first gray value coefficient and the second gray value coefficient; A coefficient difference comparison module, used for comparing the coefficient difference with a preset coefficient difference threshold; The contrast adjustment determination module is used to perform contrast adjustment on the fingerprint image data when the coefficient difference is lower than a preset coefficient difference threshold.

[0023] The technical effect of the above technical solution is: the fingerprint image data of the patient is obtained through the image extraction module, which provides a basis for subsequent processing. The area division module can accurately distinguish the fingerprint area and the non-fingerprint area, avoiding the interference of the non-fingerprint area on the fingerprint recognition result. By calculating the first gray value coefficient and the second gray value coefficient and comparing the coefficient difference between them, the quality of the fingerprint image can be further evaluated. When the coefficient difference is lower than the preset threshold, it indicates that the fingerprint image may have a problem of insufficient contrast. At this time, contrast adjustment can improve the accuracy of fingerprint recognition. The contrast adjustment judgment module triggers the contrast adjustment operation when detecting that the fingerprint image has insufficient contrast. This helps to improve the visual effect of the fingerprint image, making the fingerprint texture clearer, thereby improving the success rate of fingerprint recognition. The technical solution can automatically determine whether the fingerprint image needs contrast adjustment by calculating the gray value coefficient and comparing the coefficient difference. This adaptive processing method enables the fingerprint recognition module to cope with fingerprint images of different qualities and conditions, and improves the adaptability and robustness of the module. By optimizing the fingerprint recognition process, the recognition failure or delay caused by poor fingerprint image quality is reduced. This improves the user's experience and satisfaction when using the fingerprint recognition function. This technical solution proposes a new fingerprint image processing method, which provides new ideas and methods for the development of fingerprint recognition technology. This will help promote the application and development of fingerprint recognition technology in the fields of security authentication and identity recognition.

[0024] On the other hand, by accurately distinguishing the fingerprint area and the non-fingerprint area, and using the gray value coefficient to evaluate the image quality, the technical solution can more accurately identify the fingerprint features, thereby reducing the false recognition rate. Under the action of the contrast adjustment judgment module, when the fingerprint image quality is poor, the clarity of the fingerprint texture can be improved by adjusting the contrast, thereby improving the rejection rate and reducing the misjudgment of the real fingerprint as a mismatch. The area division module can quickly obtain the fingerprint area and the non-fingerprint area from the fingerprint image data, saving time for subsequent processing. The first gray value set acquisition module and the second gray value set acquisition module can quickly extract the gray values ​​of the fingerprint area and the non-fingerprint area, providing a basis for the calculation of the gray value coefficient. When the coefficient difference is lower than the preset coefficient difference threshold, the contrast adjustment judgment module can quickly trigger the contrast adjustment operation to ensure that the fingerprint image reaches the best recognition state in a short time. The technical solution can automatically determine whether the fingerprint image needs contrast adjustment, thereby adapting to fingerprint images of different qualities and conditions, such as dry fingers, wet fingers, shallow texture fingers, etc. Through the calculation and comparison of the gray value coefficient, the technical solution can effectively filter out noise and interference in the fingerprint image and improve the stability and reliability of fingerprint recognition. This technical solution can effectively prevent fingerprints from being cracked and misused through high-precision fingerprint recognition algorithms and image processing technology, thereby improving the security of the system. In practical applications, this technical solution can also be combined with encrypted storage and secure encrypted transmission technology to further protect the security of fingerprint data.

[0025] In summary, the technical effects of the above technical solution in terms of performance indicators are mainly reflected in the improvement of recognition accuracy, optimization of processing efficiency, enhancement of robustness and improvement of security. These technical effects make the technical solution have higher application value and competitiveness in the field of fingerprint recognition. At the same time, the technical solution improves the accuracy and adaptability of fingerprint recognition, optimizes the contrast of fingerprint images, and improves user experience through a series of processing steps and judgment logic. At the same time, the technical solution also provides new impetus and direction for the development of fingerprint recognition technology.

[0026] Specifically, the contrast adjustment determination module includes: A coefficient difference retrieving module, used for retrieving the coefficient difference between the first gray value coefficient and the second gray value coefficient when the coefficient difference is lower than a preset coefficient difference threshold; A first contrast adjustment module, configured to adjust the contrast of the fingerprint region by using a coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the first gray value coefficient; The adjusted contrast value corresponding to the fingerprint area is obtained by the following formula: Among them, D zIndicates the adjusted contrast value corresponding to the fingerprint area; D z0 Indicates the contrast value before adjustment corresponding to the fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold; A second contrast adjustment module, configured to adjust the contrast of the non-fingerprint area by using the coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the second gray value coefficient; The adjusted contrast value corresponding to the non-fingerprint area is obtained by the following formula: Among them, D f Indicates the adjusted contrast value corresponding to the non-fingerprint area; D f0 Indicates the contrast value before adjustment corresponding to the non-fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold.

[0027] The technical effect of the above technical solution is: by contrast adjustment, the overall visual effect of the fingerprint image can be improved, the fingerprint texture is clearer, and the boundary between the non-fingerprint area and the fingerprint area is clearer. This helps to improve the accuracy of subsequent fingerprint recognition. Contrast adjustment helps to highlight the fingerprint features, reduce noise and interference, and thus improve the speed and accuracy of fingerprint recognition. In particular, in the case of low-quality fingerprint images, contrast adjustment can significantly improve the recognition performance. The technical solution can automatically determine and adjust the contrast of the fingerprint image to adapt to changes in image quality caused by factors such as different lighting conditions and finger humidity. This enhances the robustness and adaptability of the fingerprint recognition system. The contrast adjustment module performs different processing on the fingerprint area and the non-fingerprint area respectively. This regional adjustment method is more flexible and can better retain important information in the fingerprint image. Through the preset coefficient difference threshold (Jcy), the user can adjust the sensitivity of the contrast adjustment according to actual needs. This configurability makes the technical solution more suitable for different application scenarios and needs. By optimizing the fingerprint recognition process, the recognition failure or delay caused by poor image quality is reduced, thereby improving the user experience and satisfaction when using the fingerprint recognition function.

[0028] On the other hand, by using the first contrast adjustment module, the contrast of the fingerprint area is adjusted by using the coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the first gray value coefficient, which can significantly improve the clarity of the fingerprint area, making the fingerprint texture clearer, which is conducive to subsequent fingerprint recognition. The second contrast adjustment module uses the same coefficient difference in combination with the second gray value coefficient to adjust the contrast of the non-fingerprint area, which helps to reduce the interference of the non-fingerprint area on fingerprint recognition and improve the accuracy of recognition. The quality of the fingerprint image after contrast adjustment is improved, and the fingerprint features are more obvious, thereby improving the accuracy of fingerprint recognition. A clear fingerprint image is conducive to the fingerprint recognition algorithm to extract fingerprint features more quickly and compare, thereby speeding up the recognition speed. The technical solution can automatically determine and adjust the contrast of the fingerprint image, adapt to the changes in image quality caused by external factors such as different lighting conditions and finger humidity, and enhance the robustness of the system. Through contrast adjustment, the technical solution can process fingerprint images of various qualities, including low-quality images, and improve the adaptability of the system. The combination of clear fingerprint images and accurate recognition algorithms can significantly reduce the number of recognition failures and improve the user experience. Fast and accurate fingerprint recognition can improve user satisfaction with the system, especially in scenarios where fingerprint recognition is frequently used. Although the technical solution itself does not directly mention security and privacy protection, clear fingerprint images and accurate recognition can help prevent the risk of personal information leakage caused by recognition errors. Accurate fingerprint recognition can ensure that only legitimate users can access the system or device, thereby enhancing the security of the system.

[0029] In summary, this technical solution optimizes the fingerprint image quality through the contrast adjustment judgment module, improves the accuracy and robustness of fingerprint recognition, enhances the flexibility and configurability of the system, and improves the user experience.

[0030] In this embodiment, the pressure data analysis unit includes: The pressure change monitoring module is used to monitor the dynamic changes of the patient's finger pressure data based on the time series of the pressure change data set, and perform fluctuation analysis on the pressure data in each time period. It can identify abnormal pressure fluctuation patterns related to Parkinson's disease, such as sudden pressure increase or decrease or unchanged pressure, and generate a pressure change report; The behavioral feature analysis module is used to extract behavioral features associated with finger movement from the finger pressure change data, such as tremor, bradykinesia, finger force change and other behavioral features. Combined with the time series of pressure changes, it can determine the reaction speed of the patient's fingers in different movements and evaluate the degree of bradykinesia. The arteriovenous recognition module is used to capture the vein map of the patient's finger based on infrared imaging equipment, and distinguish between live and non-live samples in the patient's finger pressure data in combination with liveness detection technology, ensuring that the collected data is real, not forged or tampered with biometric features, effectively preventing interference from false identities, while protecting the patient's privacy information from being leaked, and ensuring that the system can prevent the data from being affected by prostheses or other external factors; The behavior feature analysis module combines the recognition result of the arteriovenous recognition module and issues a false data abnormality alarm when it recognizes that there are non-living samples in the finger pressure data.

[0031] In this embodiment, by accurately monitoring the changes in the patient's finger pressure and identifying abnormal patterns related to Parkinson's disease, such as tremor and bradykinesia, effective support is provided for early diagnosis. At the same time, by combining arteriovenous recognition technology and liveness detection technology, the authenticity and integrity of data collection are ensured, interference from prostheses or false identities is avoided, and patient privacy is effectively protected. Through real-time monitoring, precise analysis and abnormal alarms, the reliability and accuracy of diagnosis are improved, and doctors are provided with more accurate decision-making support, further optimizing the intelligent diagnosis process of Parkinson's disease.

[0032] In this embodiment, the behavior feature analysis module specifically includes: Perform frequency domain analysis on finger pressure data to extract tremor frequency and amplitude. Determine the frequency peak based on the extraction results to identify whether the patient has tremor behavior, calculate the tremor amplitude, and determine the patient's tremor frequency. Usually, Parkinson's disease patients have a tremor frequency range of 4-8 Hz. Based on the time series of pressure changes, the time delay of each pressure fluctuation cycle is calculated, and the patient's reaction time is evaluated based on the degree of time delay to determine the patient's bradykinesia index. Parkinson's patients usually show delayed reactions and slower pressure increase and decrease rates. Analyze the fluctuation of pressure data during continuous force application, quantify the force fluctuation through standard deviation and the difference between maximum and minimum values, and determine the patient's finger force fluctuation. Unstable or excessive force changes may be a sign of motor control problems; The pressure data and behavioral characteristics of each action process are integrated in time series, the time axis is aligned, and the reaction speed and relative changes of the action are detected; Calculate the reaction time of pressure data for different finger movements (such as clenching a fist, releasing the hand, fingertip touching, etc.) to assess whether the patient has delayed behavior; Based on the time series, the changing trends of the behavioral characteristics of the fingers under different movements are quantified, and the abnormal scores of the behavioral characteristics are generated, and the patient's behavioral characteristics report is generated, including tremor amplitude, bradykinesia index, force fluctuations, etc.

[0033] In this embodiment, by performing frequency domain analysis, time series calculation, force fluctuation quantification and time series fusion on the finger pressure data, the patient's tremor behavior can be effectively identified, reaction time can be evaluated, bradykinesia can be detected and force fluctuations can be quantified, thereby providing a detailed behavioral characteristic report for the diagnosis of movement disorders such as Parkinson's disease, improving the recognition accuracy of tremor symptoms, accurately evaluating the patient's bradykinesia, quantifying the instability of finger strength, real-time monitoring of the reaction speed of different finger movements, and comprehensive analysis of the changing trends of behavioral characteristics, providing doctors with a comprehensive diagnostic basis, facilitating early detection and intervention of movement dysfunction, improving the accuracy and efficiency of diagnosis, and also providing an important reference for the personalized treatment of patients.

[0034] In this embodiment, the Parkinson's disease diagnosis unit includes: The comprehensive diagnosis module is used to compare the comprehensive abnormality score with the set threshold of the clinical diagnosis standard for Parkinson's disease. If the comprehensive score exceeds the threshold, it indicates that there may be symptoms related to Parkinson's disease, which assists doctors in making diagnostic decisions; The diagnostic result generation module is used to generate diagnostic suggestions based on the comprehensive abnormality score and patient historical data to assist doctors in determining whether patients have Parkinson's disease symptoms; The comprehensive abnormality score specifically includes: The finger pressure data, behavioral characteristics, vein map and liveness detection results are weightedly fused to generate a unified multi-dimensional feature vector, and the time series characteristics of the multi-dimensional feature vector are extracted; Classifying the fused multi-dimensional feature vectors to identify behavioral features associated with Parkinson's disease, including tremor, bradykinesia, force fluctuation, etc. At the same time, generating an abnormal score for each behavioral feature based on the classification results, including but not limited to abnormal scores for tremor, bradykinesia and force fluctuation, wherein the abnormal score is determined based on comparison with historical clinical data; By weighted summing up the abnormal scores of each behavioral feature, a comprehensive abnormal score for each patient is generated, and combined with the set threshold, it is determined whether there is a Parkinson's disease-related behavioral pattern; Based on the patient's historical diagnosis and treatment data, it is determined whether the current abnormal behavior characteristics are new symptoms or changes in existing symptoms, and targeted treatment recommendations are provided to doctors.

[0035] In this embodiment, by combining multi-dimensional data such as finger pressure, movement characteristics, vein map and liveness detection information, a comprehensive abnormality score is generated to comprehensively evaluate the patient's health status and effectively identify behavioral characteristics related to Parkinson's disease. By weighted fusion and combining the comparison of the patient's historical clinical data, the accuracy and personalization of the abnormality score are ensured. The comparison of the comprehensive abnormality score with the set threshold can promptly indicate possible Parkinson's disease symptoms, providing doctors with strong auxiliary diagnosis support, thereby improving the accuracy of diagnosis and reducing the risk of misdiagnosis and missed diagnosis.

[0036] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. Intelligent diagnosis system for Parkinson's disease based on finger pressure recognition, characterized in that: include: A biometric feature collection unit is used to collect the patient's finger pressure data based on fingerprint recognition and finger pressure sensors, and at the same time, obtain the patient's biometric data and verify and screen the patient's identity; A pressure data analysis unit, which is used to monitor dynamic pressure changes based on the patient's finger pressure data, and to determine whether the patient has Parkinson's disease-related behavioral characteristics by analyzing the dynamic pressure changes of the fingers; The Parkinson's disease diagnosis unit is used to integrate the collected biometric data, behavioral characteristics and dynamic pressure analysis data, analyze the possible symptoms of Parkinson's disease, compare them with the medical expertise database, generate diagnosis results, and provide auxiliary diagnosis suggestions.

2. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 1, characterized in that: Biometric feature collection unit, including: The fingerprint recognition module is used to obtain the patient's fingerprint image data in real time based on the fingerprint sensor, and pre-process the collected fingerprint image data to extract the patient's fingerprint features from the processed fingerprint image; The pressure sensing module is used to collect the pressure distribution on the patient's fingers through the finger pressure sensor. At the same time, each finger is labeled based on the patient's fingerprint characteristics, and a corresponding pressure change data set is generated based on each finger label.

3. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 2, characterized in that: Fingerprint recognition module, including: An image extraction module, used to extract the patient's fingerprint image data; A region division module, used for obtaining a fingerprint region and a non-fingerprint region from the fingerprint image data; A first grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixels contained in the fingerprint area as a first grayscale value set; A second grayscale value set acquisition module, used to extract grayscale values ​​corresponding to the pixel points included in the non-fingerprint area as a second grayscale value set; A first gray value coefficient acquisition module, used to acquire a first gray value coefficient by using the gray value of the pixel point included in the first gray value set; The first gray value coefficient is obtained by the following formula: Among them, J 01 represents the first gray value coefficient; n represents the number of pixels contained in the first gray value set; H 01i represents the gray value of the i-th pixel point included in the first gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H c Indicates the preset grayscale reference value; H 01bi Represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the first grayscale value set; A second gray value coefficient acquisition module, used to acquire a second gray value coefficient by using the gray value of the pixel point included in the second gray value set; The second gray value coefficient is obtained by the following formula: Among them, J 02 represents the second gray value coefficient; m represents the number of pixels included in the second gray value set; H 02i represents the gray value of the i-th pixel point included in the second gray value set; H 01z represents the grayscale intermediate value corresponding to the first grayscale value set; H 02z represents the grayscale intermediate value corresponding to the second grayscale value set; H 02bi represents the grayscale standard deviation of the i-th pixel and its adjacent pixels included in the second grayscale value set; A coefficient difference acquisition module, used for comparing the first gray value coefficient with the second gray value coefficient to acquire the coefficient difference between the first gray value coefficient and the second gray value coefficient; A coefficient difference comparison module, used for comparing the coefficient difference with a preset coefficient difference threshold; The contrast adjustment determination module is used to perform contrast adjustment on the fingerprint image data when the coefficient difference is lower than a preset coefficient difference threshold.

4. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 3, characterized in that: The contrast adjustment determination module comprises: A coefficient difference retrieving module, used for retrieving the coefficient difference between the first gray value coefficient and the second gray value coefficient when the coefficient difference is lower than a preset coefficient difference threshold; A first contrast adjustment module, configured to adjust the contrast of the fingerprint region by using a coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the first gray value coefficient; The adjusted contrast value corresponding to the fingerprint area is obtained by the following formula: Among them, D z Indicates the adjusted contrast value corresponding to the fingerprint area; D z0 Indicates the contrast value before adjustment corresponding to the fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold; A second contrast adjustment module, configured to adjust the contrast of the non-fingerprint area by using the coefficient difference between the first gray value coefficient and the second gray value coefficient in combination with the second gray value coefficient; The adjusted contrast value corresponding to the non-fingerprint area is obtained by the following formula: Among them, D f Indicates the adjusted contrast value corresponding to the non-fingerprint area; D f0 Indicates the contrast value before adjustment corresponding to the non-fingerprint area; J 01 Represents the first gray value coefficient; J 02 Represents the second gray value coefficient; J cy Indicates the preset coefficient difference threshold.

5. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 2, characterized in that: The biometric feature collection unit also includes: The identity verification module is used to compare the extracted fingerprint features with the patient identity data in the pre-stored patient database to verify the patient's identity information. If the verification fails, an identity abnormality warning is issued and the monitoring operation is terminated; The file management module is used to associate the extracted fingerprint feature data with the patient's identity information, construct the patient's personal file data, and manage and store the acquired patient's personal file data; The data storage module is used to store the collected patient biometric data, health records and diagnostic data, update the patient's personal file data after each visit, and synchronize the updated personal file data to the database in real time.

6. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 5, characterized in that: Pressure data analysis unit, including: The pressure change monitoring module is used to monitor the dynamic changes of the patient's finger pressure data based on the time series of the pressure change data set, perform fluctuation analysis on the pressure data in each time period, and generate a pressure change report; The behavior feature analysis module is used to extract the behavior features associated with finger movement from the finger pressure change data, and combine the time series of pressure changes to determine the reaction speed of the patient's fingers during different movements and evaluate the degree of bradykinesia; The arteriovenous recognition module is used to capture the vein map of the patient's finger based on infrared imaging equipment, and distinguish between live and non-live samples in the patient's finger pressure data in combination with liveness detection technology; The behavior feature analysis module combines the recognition result of the arteriovenous recognition module and issues a false data abnormality alarm when it recognizes that there are non-living samples in the finger pressure data.

7. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 6, characterized in that: Behavior feature analysis module, including: Perform frequency domain analysis on finger pressure data to extract tremor frequency and amplitude, determine the frequency peak based on the extraction results, identify whether the patient has tremor behavior, calculate the tremor amplitude, and determine the patient's tremor frequency; Based on the time series of pressure changes, the time delay of each pressure fluctuation cycle is calculated, the patient's reaction time is evaluated based on the degree of time delay, and the patient's bradykinesia index is determined; Analyze the fluctuation of pressure data during continuous force application, quantify the force fluctuation through standard deviation, maximum value and minimum value difference, and determine the patient's finger force fluctuation; The pressure data and behavioral characteristics of each action process are integrated in time series, the time axis is aligned, and the reaction speed and relative changes of the action are detected; Calculate the reaction time of pressure data for different finger movements to assess whether the patient has delayed behavior; Based on the time series, the changing trends of the behavioral characteristics of the fingers under different movements are quantified, and the abnormal scores of the behavioral characteristics are generated, and the patient's behavioral characteristics report is generated.

8. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 7, characterized in that: Parkinson's disease diagnostic unit, including: The comprehensive diagnosis module is used to compare the comprehensive abnormality score with the set threshold of the clinical diagnosis standard for Parkinson's disease. If the comprehensive score exceeds the threshold, it indicates that there may be symptoms related to Parkinson's disease, which assists doctors in making diagnostic decisions; The diagnosis result generation module is used to generate diagnostic suggestions based on the comprehensive abnormality score and patient historical data to assist doctors in determining whether the patient has symptoms of Parkinson's disease.

9. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 8, characterized in that: The comprehensive abnormality score specifically includes: The finger pressure data, behavioral characteristics, vein map and liveness detection results are weightedly fused to generate a unified multi-dimensional feature vector, and the time series characteristics of the multi-dimensional feature vector are extracted; Classify the fused multi-dimensional feature vectors to identify behavioral features associated with Parkinson's disease, and generate an abnormality score for each behavioral feature based on the classification results, wherein the abnormality score is determined based on a comparison with historical clinical data; By weighted summing up the abnormal scores of each behavioral feature, a comprehensive abnormal score for each patient is generated, and combined with the set threshold, it is determined whether there is a Parkinson's disease-related behavioral pattern; And based on the patient's historical diagnosis and treatment data, determine whether the current abnormal behavior characteristics are new symptoms or changes in existing symptoms.

10. The Parkinson's disease intelligent diagnosis system based on finger pressure recognition as claimed in claim 9, characterized in that: The Parkinson's disease diagnosis unit also includes: feeding back the patient's biometric data and diagnosis results to users and medical personnel through a human-computer interaction interface, generating a health report for the patient, and providing the patient with corresponding treatment recommendations, lifestyle adjustment plans and further medical examination guidance based on the diagnosis results and auxiliary diagnosis suggestions.

Images