Disease prevention and management system based on image recognition and big data

Abstract

A disease prevention and management system based on image recognition and big data, specifically relates to the technical field of disease prevention and management of image recognition. The scheme includes an image acquisition module for acquiring image information of the current patient and converting the acquired analog signal into a digital signal. The big data processing module is used to adopt the distributed storage and parallel processing technology of big data to extract data information of the same gender and age group, and to perform disease prediction evaluation according to the normal interval and severe interval of the disease prediction severity score PF of the same gender and age group, and collect and manage after the evaluation. The feature extraction and prediction module calculates and outputs the preliminary evaluation value QT of the image feature, the abnormal area feature value YT, and the disease prediction severity score PF in sequence. The feedback display module is used to display the severity of the disease prediction. The present invention can more accurately evaluate the health status of the patient and provide a scientific basis for the prevention and management of the disease.

Description

Technical Field

[0001] The present invention relates to the technical field of disease prevention management based on image recognition, and in particular to a disease prevention management system based on image recognition and big data. Background Art

[0002] With the continuous advancement of medical technology and the advent of the big data era, disease prevention and management are gradually developing in the direction of intelligence and precision. The disease prevention and management system based on image recognition and big data is an important product of this trend. The system obtains the patient's image information through image acquisition equipment, and uses advanced image processing and big data analysis technology to monitor and warn the patient's health status in real time.

[0003] However, existing image recognition technology cannot accurately extract key feature information from images, resulting in inaccurate subsequent disease prediction results. When processing massive image data and patient information, existing big data technology faces problems of low processing efficiency and slow response speed, which will limit the real-time and accuracy of the system, thereby affecting the prevention and management of the disease. In addition, existing prediction models are often trained based on large-scale data sets, but cannot fully consider the individual differences of patients, resulting in a lack of personalization in the prediction results, which will reduce the system's prediction accuracy and practicality, and cannot adjust the image acquisition and analysis strategies according to the prediction results, which will also limit the system's self-optimization and continuous improvement capabilities.

[0004] Therefore, those skilled in the art have provided a disease prevention and management system based on image recognition and big data to solve the problems raised in the above background technology. Summary of the invention

[0005] The technical problem solved by the present invention is to provide a disease prevention and management system based on image recognition and big data to realize the personalization of prediction models, improve the accuracy of image feature extraction, enhance the efficiency of big data processing, and establish a closed-loop feedback mechanism.

[0006] To solve the above problems, the present invention provides the following technical solutions: Disease prevention and management system based on image recognition and big data, including: An image acquisition module is used to acquire the image information of the current patient and convert the acquired analog signal into a digital signal; A big data processing module is used to extract data information of the same gender and age group using distributed storage and parallel processing technology of big data, and to perform disease prediction evaluation based on the normal range and severe range of the disease prediction severity score PF of the same gender and age group, and to collect and manage the data after the evaluation; A feature extraction and prediction module, which is used to calculate and output a preliminary evaluation value QT of image features, a feature value YT of abnormal regions, and a disease prediction severity score PF in sequence based on the information transmitted by the image acquisition module and the big data processing module; A feedback display module is used to display the severity of the disease prediction under the cooperation of the big data processing module and the output of the disease prediction severity score PF result value; The feature extraction and prediction module includes a unit for extracting and quantifying image features, a unit for evaluating the strength of abnormal region features, and a unit for providing disease risk assessment.

[0007] Further: The calculation formula of the extraction and quantification image feature unit is as follows: ; in: QT is the preliminary evaluation value of image features; R is the red mean value; G is the green average value; B is the blue average value; The red mean value R, green mean value G and blue mean value B respectively identify the average degree of the red, green and blue primary color components of all pixels in the image; SD is the mean value of hue; BH is the average saturation value; L is the average brightness; The average hue value SD, the average saturation value BH and the average brightness value L are the average degrees of hue, saturation and brightness of the recognized image after the image is converted into the HSL color space; An indicator used to measure the degree of color dispersion of an image, and the more dispersed the color, the larger the value; A measure of the relative differences in image hue and saturation, and adjustments affected by brightness.

[0008] Further: The calculation formula for evaluating the characteristic strength unit of the abnormal area is as follows: ; in: YT is the characteristic value of the abnormal area; YM is the area of ​​the abnormal region, and YM reflects the size of the area where the abnormality exists in the image; ZM is the total area of ​​the image, and ZM reflects the size of the area recognized by the image as a whole; The calculation formula of the abnormal area YM is: YM = number of pixels × area of ​​each pixel, where the area of ​​each pixel is calculated by dividing the total image area ZM by the total number of pixels; Yvar is abnormal red variance; YG var is abnormal green variance; YB var is the abnormal blue variance; Abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var They are the variances of the three primary colors red, green, and blue within the abnormal region area YM; Used to measure the uniformity of color distribution and perform color analysis on abnormal areas.

[0009] Further: The calculation formula for providing the disease risk assessment unit is as follows: ; in: PF is the disease prediction severity score; NL is the current patient age; NL med is the reference age, NL med Reflects the median age of people of the same gender and age group based on big data; NL range For age range, NL range Reflects the degree of age differences within the same gender and age group; If the current patient's age NL is within the same gender and age range of 0-30, then NL range =30; If the current patient's age NL is within the same gender and age range of 31-55, then NL range =24; If the current patient's age NL is within the same gender and age range of 56-75, then NL range =19; If the current patient's age NL is within the same gender and age range of 76-100, then NL range =24; S is the sex factor, which reflects the influence of sex factors on the predicted severity of the disease; If the proportion of males in the disease is high, the S value for males is 1.2, while the S value for females is 0.8; If the proportion of women in the disease is high, the S value for female gender is 1.2, while the S value for male gender is 0.8; Used to reflect the impact of age factors on the predicted severity of the disease.

[0010] Further: Based on the disease prediction severity score PF, and the normal range and severe range of the same gender and age group extracted from big data, the prediction analysis is as follows: If the disease prediction severity score PF is within the normal range, it means that the image characteristics of the current patient are similar to those of the normal population and the disease risk is low; If the disease prediction severity score PF is in the severe range or above the range, it means that the image features of the current patient have abnormal changes related to known diseases, and the patient should be reminded to undergo further examination and treatment; Among them, the normal interval and the severe interval are the disease prediction severity score PF intervals of normal and abnormal persons respectively displayed by image recognition.

[0011] Further: the disease prediction severity score PF, the current patient age NL and the current patient's gender will be automatically collected and summarized into big data and divided into normal range, severe range, reference age NL med , Age range NL range And gender factor S adds the basis for big data.

[0012] Further: the equipment used in the image acquisition module includes a high-resolution camera and an image acquisition card; The equipment used by the big data processing module includes big data processing equipment; The equipment used in the feature extraction prediction module includes a server and a high-performance computer; The equipment used by the feedback display module includes a display device.

[0013] Further: The abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var The calculation formulas are as follows: ; ; ; in: N1 is an abnormal red pixel; Y i is the i-th red pixel value in the abnormal area; YR is the average value of abnormal red pixels; N2 is an abnormal green pixel; YG i is the i-th green pixel value in the abnormal area; YG is the average value of abnormal green pixels; N3 is an abnormal blue pixel; YB i is the i-th blue pixel value in the abnormal area; YB is the average value of abnormal blue pixels.

[0014] The effects of the above solution are as follows: 1. The present invention firstly improves the accuracy of image feature extraction by optimizing the image processing algorithm and introducing advanced feature extraction technology, which enables the system to more accurately identify key information in the image and provide reliable data support for subsequent disease prediction. Secondly, the use of distributed storage and parallel processing technology can improve the efficiency of big data processing, which enables the system to process massive image data and patient information faster and improve the real-time performance and accuracy of the system.

[0015] 2. The present invention combines the individual differences of patients of the same gender and age group, including the reference age NL med and age range NL range , personalize the prediction model, which enables the system to more accurately predict the current health status of the patient, thereby improving the accuracy and practicality of the prediction.

[0016] 3. The present invention provides the result P of the disease risk assessment unit, i.e., the disease prediction severity score PF, and then establishes a closed-loop feedback mechanism. When the disease prediction severity score PF is within the severe interval and is higher than the interval, the system can automatically adjust the image acquisition and analysis strategy to improve the accuracy of subsequent analysis. At the same time, the disease prediction severity score PF and the patient's age and gender information are summarized in the big data storage for updating and sorting, which can provide data support for the continuous optimization and improvement of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of the method of the disease prevention management system; Figure 2 This is a schematic diagram of the overall structure of the disease prevention and management system; Figure 3 It is a structural schematic diagram of the feature extraction prediction module in the present invention. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely introduced below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0019] For example, see Figure 1-3 , a disease prevention and management system based on image recognition and big data, including: An image acquisition module is used to acquire the image information of the current patient and convert the acquired analog signal into a digital signal; A big data processing module is used to extract data information of the same gender and age group using distributed storage and parallel processing technology of big data, and to perform disease prediction evaluation based on the normal range and severe range of the disease prediction severity score PF of the same gender and age group, and to collect and manage the data after the evaluation; The feature extraction and prediction module is used to calculate and output the preliminary evaluation value QT of the image feature, the feature value YT of the abnormal area, and the disease prediction severity score PF in sequence based on the information transmitted by the image acquisition module and the big data processing module; A feedback display module is used to display the severity of the disease prediction under the cooperation of the big data processing module and the output of the disease prediction severity score PF result value; Among them, the feature extraction prediction module includes a unit for extracting and quantifying image features, a unit for evaluating the feature strength of abnormal regions, and a unit for providing disease risk assessment; The equipment used in the image acquisition module includes a high-resolution camera and an image acquisition card; The equipment used in the big data processing module includes big data processing equipment; The equipment used in the feature extraction prediction module includes servers and high-performance computers; The equipment used in the feedback display module includes a display device.

[0020] This embodiment develops a disease prevention management system based on image recognition and big data, which realizes the collection, processing, feature extraction, disease prediction severity assessment, feedback and optimization functions of patient image information through a series of method steps, modules and units and the equipment used, and provides strong support for disease prevention and management; In this embodiment, in the disease prevention and management system based on image recognition and big data, since it is necessary to process a large amount of current and historical patient image information, the system adopts distributed storage technology to store the data in a dispersed manner on multiple nodes to improve data access speed and system scalability. When processing big data, in order to improve processing efficiency, the system adopts parallel processing technology, which means splitting the big data task into multiple subtasks and executing these subtasks simultaneously on multiple processors and computing nodes. This can significantly shorten the processing time and improve the real-time performance and response speed of the system.

[0021] See also Figure 1-3 , the calculation formula for extracting and quantifying image feature units is as follows: ; in: QT is the preliminary evaluation value of image features; R is the red mean value; G is the green average value; B is the blue average value; The red mean value R, green mean value G and blue mean value B respectively identify the average degree of the red, green and blue primary color components of all pixels in the image; SD is the mean value of hue; BH is the average saturation value; L is the average brightness; The average hue value SD, the average saturation value BH and the average brightness value L are the average degrees of hue, saturation and brightness of the recognized image after the image is converted into the HSL color space; An indicator used to measure the degree of color dispersion of an image, and the more dispersed the color, the larger the value; A measure of the relative differences in image hue and saturation, and adjustments affected by brightness.

[0022] This embodiment first The calculation part is to measure the color richness and brightness contrast of the image. This expression reflects the dispersion of colors in the image by calculating the difference between the square of the sum of the three primary colors and the sum of their respective squares. When the color distribution in the image is more uniform, the difference will be relatively small, and when the color distribution in the image is more dispersed, the difference will be relatively large. This calculation method can capture the overall distribution of image colors. As an important part of the preliminary evaluation value QT of image features, this calculation part provides information about the color richness and brightness contrast of the image, which is helpful for subsequent disease prediction analysis; In the calculation part, the hue average value SD and the saturation average value BH are important parameters that reflect the color characteristics of the image. By dividing the difference between hue and saturation by the brightness average value L, a relative difference value affected by brightness can be obtained. This calculation method helps to eliminate the influence of brightness on the comparison of hue and saturation, thereby more accurately reflecting the characteristics of image color. This calculation part is combined with the color dispersion index to form the preliminary evaluation value QT of image features, which provides information about the relative difference between image hue and saturation, and helps to enhance the accuracy of disease prediction; In this algorithm unit, the red average value R, green average value G and blue average value B directly reflect the intensity distribution of the image in the three color channels of red, green and blue, which is crucial for identifying color changes in images. By calculating the statistical characteristics of these channel values, the color dispersion of the image can be captured, thereby more accurately extracting the key features of the image; The hue average value SD, saturation average value BH and brightness average value L provide another way to represent the image in the color space. In particular, the brightness average value L is of great significance for measuring the brightness contrast of the image. By comparing the brightness values ​​of different areas, the light and dark changes in the image can be identified and the description of the image features can be further enriched. The calculation process of these parameters in the image feature extraction and quantification unit is relatively simple and efficient, enabling the system to quickly respond and extract image features in real-time applications. The resulting image feature preliminary evaluation value QT of the image feature extraction and quantification unit serves as a preliminary representation of the image features, providing important input data for subsequent processing.

[0023] At the same time, the preliminary evaluation value QT of image features, as a standardized representation of image features, also facilitates data exchange and sharing between different systems, thereby promoting cross-platform cooperation and knowledge sharing.

[0024] See also Figure 1-3 , the calculation formula for evaluating the characteristic intensity unit of the abnormal area is as follows: ; in: YT is the characteristic value of the abnormal area; YM is the area of ​​the abnormal region, and YM reflects the size of the area where the abnormality exists in the image; ZM is the total area of ​​the image, and ZM reflects the size of the area recognized by the image as a whole; The calculation formula of the abnormal area YM is: YM = number of pixels × area of ​​each pixel, where the area of ​​each pixel is calculated by dividing the total image area ZM by the total number of pixels; Y var is abnormal red variance; YG var is abnormal green variance; YB var is the abnormal blue variance; Abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var They are the variances of the three primary colors red, green, and blue within the abnormal region area YM; Used to measure the uniformity of color distribution and perform color analysis on abnormal areas; Abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var The calculation formulas are as follows: ; ; ; in: N1 is an abnormal red pixel; Y i is the i-th red pixel value in the abnormal area; YR is the average value of abnormal red pixels; N2 is an abnormal green pixel; YG i is the i-th green pixel value in the abnormal area; YG is the average value of abnormal green pixels; N3 is an abnormal blue pixel; YB i is the i-th blue pixel value in the abnormal area YB is the average value of abnormal blue pixels.

[0025] This embodiment first The calculation part combines the preliminary calculation result of the image features with the abnormal area area YM to evaluate the feature strength of the abnormal area. By multiplying the preliminary evaluation value QT of the image features with the ratio of the abnormal area area YM to the total image area ZM, a value reflecting the feature strength of the abnormal area can be obtained. This calculation method takes into account the proportion of the abnormal area in the entire image, which helps to more accurately evaluate the feature strength of the abnormal area. As an important part of the abnormal area feature recalculation result abnormal area feature value YT, this calculation part provides information about the feature strength of the abnormal area and provides key data for subsequent disease prediction; The calculation part is used to measure the uniformity of color distribution and as a supplementary indicator of the characteristic strength of abnormal areas, where the abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var It reflects the uniformity of the color distribution in the image. By adding these variances, we can get a comprehensive index to measure the uniformity of color distribution. This calculation method helps to capture the subtle changes in color distribution in the image, so as to more comprehensively evaluate the characteristic strength of the abnormal area. The calculation parts are combined to form the abnormal region feature and then calculate the abnormal region feature value YT. The calculation part provides information about the uniformity of color distribution, which helps to enhance the accuracy and reliability of disease prediction; In this algorithm unit, the area of ​​the abnormal region YM directly reflects the area occupied by the abnormal region in the image, which is of great significance for identifying potential lesion areas. By comparing the area sizes of different regions, the severity of the lesion can be preliminarily judged. Before obtaining the area of ​​the abnormal region YM, the collected image information needs to be processed, and the image is divided into different regions using the image segmentation algorithm. These regions will correspond to different tissues, organs and lesion areas, and in the segmented image, the abnormal regions related to the disease are identified, and the identification of abnormalities will be achieved by comparing the characteristics of each region with the threshold preset by the system. Finally, for the identified abnormal region, the abnormal red variance YR is calculated using "YM=number of pixels×area of ​​each pixel". var , abnormal green variance YG var and abnormal blue variance YB var It measures the degree of color change of the abnormal area in the three color channels of red, green and blue. The changes in these parameters can reflect the color difference between the lesion area and the normal area, so as to more accurately identify the abnormal area. var , abnormal green variance YG var and abnormal blue variance YB var The uniformity of color distribution mainly reflects the stability and consistency of color in the abnormal area, which helps to identify abnormal pigmentation and congestion in the lesion. In addition, the unit for evaluating the characteristic intensity of the abnormal area can also predict the scope of the disease by identifying the size of the abnormal area. The abnormal region feature value YT of the abnormal region feature strength evaluation unit comprehensively considers the image feature preliminary evaluation value QT and the feature strength of the abnormal region. By combining these two pieces of information, a more comprehensive and accurate disease prediction model can be constructed. In addition, the abnormal region feature value YT can also be input as a feature vector into the machine learning algorithm for training and optimization, thereby further improving the accuracy and generalization ability of disease prediction; The abnormal region characteristic value YT of the abnormal region characteristic intensity unit not only reflects the prediction result of the disease, but also can be used as one of the important indicators for system performance evaluation. By monitoring the changes in the abnormal region characteristic value YT, problems in the system can be discovered in time and optimized and improved. At the same time, the abnormal region characteristic value YT can also be used as a feedback signal to adjust the image acquisition and analysis strategy, thereby improving the overall performance and reliability of the system.

[0026] See also Figure 1-3 , the calculation formula for the disease risk assessment unit is as follows: ; in: PF is the disease prediction severity score; NL is the current patient age; NL med is the reference age, NL med Reflects the median age of people of the same gender and age group based on big data; NL range For age range, NL range Reflects the degree of age differences within the same gender and age group; If the current patient's age NL is within the same gender and age range of 0-30, then NL range =30; If the current patient's age NL is within the same gender and age range of 31-55, then NL range =24; If the current patient's age NL is within the same gender and age range of 56-75, then NL range =19; If the current patient's age NL is within the same gender and age range of 76-100, then NL range =24; S is the sex factor, which reflects the influence of sex factors on the predicted severity of the disease; If the proportion of males in the disease is high, the S value for males is 1.2, while the S value for females is 0.8; If the proportion of women in the disease is high, the S value for female gender is 1.2, while the S value for male gender is 0.8; Used to reflect the impact of age factors on the predicted severity of the disease.

[0027] This embodiment first The calculation part is to calculate the abnormal area characteristics and the age adjustment factor Multiplying them together can give a disease prediction severity score PF that takes age factors into account, where NL is the current patient age, NL med is the reference age, NL range This calculation method helps to eliminate the impact of age differences on disease prediction and make the prediction results more accurate. As an important part of the disease prediction severity score PF, this calculation part provides information about the impact of age factors on disease prediction, which helps to enhance the pertinence and practicality of the prediction results. The above-mentioned age range NL for all patients of the same gender and the same age group whose current age NL is 0-30, 31-55, 56-75, 76-100 respectively range The calculation settings need to follow the update of big data and be adjusted flexibly; The gender factor S is a coefficient set according to gender, which is used to adjust the disease prediction severity score PF. In this example, this setting method takes into account the impact of gender differences on disease risk, making the prediction results more in line with the actual situation. It is combined with other calculation parts to form the disease prediction severity score PF. The introduction of the gender factor S helps to enhance the gender specificity of the prediction results and improve the accuracy and reliability of the prediction. The gender factor S is adjusted in real time for the current patient according to the update of big data. When the results of the big data update show that men account for a high proportion of the disease, the S value of male gender is 1.2. In the calculation part, we ensure The calculation result of , and then add the influence of high male proportion on this result, if the current patient is female, S value is 0.8, and In the calculation part, we ensure The calculation result weakens the impact of the low proportion of women on this basis. When the results of the big data update show that women account for a high proportion of the disease, the calculation impact is the same as above according to different values; The disease risk assessment unit of this algorithm takes into account the impact of different patients' physiological and psychological differences on disease prediction by introducing gender factor S and age adjustment items, which enables the system to make more accurate predictions based on individual differences of different patients. In addition, the disease risk assessment unit can also be adjusted and optimized according to the patient's specific medical history and symptom information, thereby further improving the personalization of the prediction; Providing the result of the disease risk assessment unit, the disease prediction severity score PF, can provide an important basis for formulating prevention and treatment plans, and can judge the patient's disease severity according to the size of the disease prediction severity score PF, and formulate corresponding preventive measures or treatment plans accordingly. In addition, the disease prediction severity score PF can also be used as one of the monitoring indicators of the patient's health status to evaluate the treatment effect and changes in the condition.

[0028] For example 2, please refer to Figure 1-3 The prediction analysis based on the disease prediction severity score PF and the normal range and severe range of the same gender and age group extracted from big data is as follows: If the disease prediction severity score PF is within the normal range, it means that the image characteristics of the current patient are similar to those of the normal population and the disease risk is low; If the disease prediction severity score PF is in the severe range or above the range, it means that the image features of the current patient have abnormal changes related to known diseases, and the patient should be reminded to undergo further examination and treatment; Among them, the normal interval and the severe interval are the disease prediction severity score PF intervals of normal and abnormal persons respectively displayed by image recognition; The disease prediction severity score PF, the current patient age NL, and the current patient's gender will be automatically collected and summarized into big data and divided into normal range, severe range, reference age NL med , Age range NL range And gender factor S adds the basis for big data.

[0029] In this embodiment, when the disease prediction severity score PF is within the severe interval, and above the interval, the system can automatically adjust the image acquisition parameters to improve the accuracy of subsequent analysis. This optimization mechanism can make the system more intelligent and adaptive. Through the cyclic influence mechanism, the disease risk assessment unit can guide the image feature extraction and quantification unit to optimize the acquisition and processing of image features, thereby improving the performance of the entire system. This optimization is not only reflected in the accuracy of disease prediction, but also in the stability and robustness of the system. In addition, the cyclic influence mechanism of the disease risk assessment unit on the image feature extraction and quantification unit provides the possibility for continuous improvement of the system. By continuously collecting and analyzing data, the system can gradually optimize the algorithm and parameter settings to adapt to the ever-changing data environment and clinical needs. Combining the unit for extracting and quantifying image features, the unit for evaluating the strength of abnormal area features, and the unit for providing disease risk assessment, we can clearly see the beneficial effects of image recognition and big data in disease prevention and prediction. This combination not only improves the accuracy of disease prediction, but also enhances the pertinence and effectiveness of preventive measures. The following is a detailed description of the beneficial effects brought about by this combination: The unit for extracting and quantifying image features preliminarily extracts the key features of the image by calculating the color richness and brightness contrast of the image. This process is similar to a "physical examination" of the image, which can quickly capture abnormal signals in the image. In disease prevention and prediction, this means that the system can automatically screen out images suspected of having health problems, providing important clues for subsequent analysis. Combined with the normal range and severe range in the big data storage, the system can further determine the degree of abnormality of the image features. This comparative analysis based on big data greatly improves the accuracy of disease prediction; On the basis of the unit for extracting and quantifying image features, the unit for evaluating the feature strength of abnormal regions performs a deeper feature extraction and analysis of abnormal regions. By calculating the color variance parameters of abnormal regions, the system can more accurately evaluate the feature strength of abnormal regions, thereby further narrowing the scope of disease prediction. By comparing with the severe interval of abnormal image feature values ​​in the big data storage, the system can more accurately determine the degree of association between abnormal regions and known diseases, which helps to quickly locate potential health problems and take corresponding preventive measures. The disease risk assessment unit combines image features with the patient's age and gender information to form a comprehensive disease prediction model. This model not only takes into account the image features themselves, but also fully considers the impact of individual differences on disease prediction, thereby improving the accuracy and reliability of the prediction. Based on the results of the disease risk assessment unit, the system can provide patients with personalized prevention prediction suggestions. This targeted prevention prediction helps reduce the incidence and severity of the disease. With the continuous addition of new image data and the continuous updating of disease knowledge, the normal range and severe range in the big data storage will also be continuously optimized and improved, which means that the disease prediction ability of the system will continue to improve over time. The disease prevention and prediction system combining image recognition and big data is a process of continuous iteration and upgrading. Through continuous optimization of algorithms, improvement of image recognition accuracy and expansion of big data scale, the system can continuously adapt to new disease challenges and individual needs. In summary, image recognition and big data combined with the extraction and quantification of image feature units, evaluation of abnormal area feature intensity units and provision of disease risk assessment units have shown significant beneficial effects in disease prevention and prediction. This combination not only improves the accuracy, pertinence and effectiveness of disease prediction, but also provides a solid foundation for continuous optimization and updating of the system. With the continuous advancement of technology and the continuous expansion of application scenarios, this combination is expected to play a more important role in the field of disease prevention.

[0030] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined by the claims.

Claims

1. A disease prevention and management system based on image recognition and big data, characterized in that: include: An image acquisition module is used to acquire the image information of the current patient and convert the acquired analog signal into a digital signal; A big data processing module is used to extract data information of the same gender and age group using distributed storage and parallel processing technology of big data, and to perform disease prediction evaluation based on the normal range and severe range of the disease prediction severity score PF of the same gender and age group, and to collect and manage the data after the evaluation; A feature extraction and prediction module, which is used to calculate and output a preliminary evaluation value QT of image features, a feature value YT of abnormal regions, and a disease prediction severity score PF in sequence based on the information transmitted by the image acquisition module and the big data processing module; A feedback display module is used to display the severity of the disease prediction under the cooperation of the big data processing module and the output of the disease prediction severity score PF result value; The feature extraction and prediction module includes a unit for extracting and quantifying image features, a unit for evaluating the strength of abnormal region features, and a unit for providing disease risk assessment.

2. The disease prevention and management system based on image recognition and big data according to claim 1 is characterized by: The calculation formula of the extraction and quantification image feature unit is as follows: ; in: QT is the preliminary evaluation value of image features; R is the red mean value; G is the green average value; B is the blue average value; The red mean value R, green mean value G and blue mean value B are the average degrees of the red, green and blue primary color components of all pixels in the recognition image respectively; SD is the mean value of hue; BH is the average saturation value; L is the average brightness; The average hue value SD, the average saturation value BH and the average brightness value L are the average degrees of hue, saturation and brightness of the recognized image after the image is converted into the HSL color space; An indicator used to measure the degree of color dispersion of an image, and the more dispersed the color, the larger the value; A measure of the relative differences in image hue and saturation, and adjustments affected by brightness.

3. The disease prevention and management system based on image recognition and big data according to claim 2 is characterized in that: The calculation formula for evaluating the characteristic intensity unit of the abnormal area is as follows: ; in: YT is the characteristic value of the abnormal area; YM is the area of ​​the abnormal region, and YM reflects the size of the area where the abnormality exists in the image; ZM is the total area of ​​the image, and ZM reflects the size of the area recognized by the image as a whole; The calculation formula of the abnormal area YM is: YM = number of pixels × area of ​​each pixel, where the area of ​​each pixel is calculated by dividing the total image area ZM by the total number of pixels; Y var is abnormal red variance; YG var is the abnormal green variance; YB var is the abnormal blue variance; Abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var They are the variances of the three primary colors red, green, and blue within the abnormal region area YM; Used to measure the uniformity of color distribution and perform color analysis on abnormal areas.

4. The disease prevention and management system based on image recognition and big data according to claim 3 is characterized by: The calculation formula for providing the disease risk assessment unit is as follows: ； in: PF is the disease prediction severity score; NL is the current patient age; NL med is the reference age, NL med Reflects the median age of the same gender and age group based on big data; NL range For age range, NL range Reflects the degree of age differences within the same gender and age group; If the current patient's age NL is within the same gender and age range of 0-30, then NL range =30; If the current patient's age NL is within the same gender and age range of 31-55, then NL range =24; If the current patient's age NL is within the same gender and age range of 56-75, then NL range =19; If the current patient's age NL is within the same gender and age range of 76-100, then NL range =24; S is the sex factor, which reflects the influence of sex factors on the predicted severity of the disease; If the proportion of males in the disease is high, the S value for males is 1.2, while the S value for females is 0.8; If the proportion of women in the disease is high, the S value for female gender is 1.2, while the S value for male gender is 0.8; Used to reflect the impact of age factors on the predicted severity of the disease.

5. The disease prevention and management system based on image recognition and big data according to claim 4 is characterized by: The prediction analysis based on the disease prediction severity score PF and the normal range and severe range of the same gender and age group extracted from big data is as follows: If the disease prediction severity score PF is within the normal range, it means that the image characteristics of the current patient are similar to those of the normal population and the disease risk is low; If the disease prediction severity score PF is in the severe range or above the range, it means that the image features of the current patient have abnormal changes related to known diseases, and the patient should be reminded to undergo further examination and treatment; Among them, the normal interval and the severe interval are the disease prediction severity score PF intervals of normal and abnormal persons respectively displayed by image recognition.

6. The disease prevention and management system based on image recognition and big data according to claim 5 is characterized by: The disease prediction severity score PF, the current patient age NL, and the current patient's gender will be automatically collected and summarized into big data and divided into normal range, severe range, reference age NL, and med , Age range NL range And gender factor S adds the basis for big data.

7. The disease prevention and management system based on image recognition and big data according to claim 1 is characterized by: The equipment used in the image acquisition module includes a high-resolution camera and an image acquisition card; The equipment used by the big data processing module includes big data processing equipment; The equipment used in the feature extraction prediction module includes a server and a high-performance computer; The equipment used by the feedback display module includes a display device.

8. The disease prevention and management system based on image recognition and big data according to claim 3 is characterized by: The abnormal red variance YR var , abnormal green variance YG var and abnormal blue variance YB var The calculation formulas are as follows: ; ; ; in: N1 is an abnormal red pixel; Y i is the i-th red pixel value in the abnormal area; YR is the average value of abnormal red pixels; N2 is an abnormal green pixel; YG i is the i-th green pixel value in the abnormal area; YG is the average value of abnormal green pixels; N3 is an abnormal blue pixel; YB i is the i-th blue pixel value in the abnormal area; YB is the average value of abnormal blue pixels.

Images