Medical data intelligent analysis method and system based on infrared image

Abstract

The invention provides a medical data intelligent analysis method and system based on an infrared image. The method comprises the following steps: acquiring first infrared image data; determining a detection target and a detection scheme of a second infrared detection terminal according to the first infrared image data and an infrared detection model; controlling a second infrared detection terminal to detect according to the detection target and the detection scheme to obtain second infrared image data; analyzing the second infrared image by using an infrared image analysis model to obtain an infrared image analysis result; determining an auxiliary image detection scheme according to the detection target and the body data of the detected object; detecting the detected object according to the auxiliary image detection scheme to obtain auxiliary image data; analyzing the auxiliary image data by using the auxiliary image analysis model to obtain an auxiliary image analysis result; and obtaining a comprehensive analysis result according to the infrared image analysis result and the auxiliary image analysis result. According to the scheme, the detection accuracy and efficiency can be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and particularly to an intelligent analysis method and system for medical data based on infrared images. Background Art

[0002] Infrared thermal imaging technology provides an effective scientific basis for modern medicine. This technology is a fast, simple, complete and non-invasive detection method, bringing many conveniences to health examinations. Through an infrared thermal imaging system, infrared radiation emitted by the human body can be collected, and these signals are converted into digital signals to form a pseudo-color thermal map. Professional physicians analyze these thermal maps and calorific values based on their clinical experience to determine abnormal parts, disease nature and lesion degree of the human body, and generate a health report. However, the existing medical data analysis methods based on infrared images have low detection efficiency and inaccurate detection results. Summary of the Invention

[0003] Based on the above problems, the present invention proposes an intelligent analysis method and system for medical data based on infrared images. Through an intelligent analysis method and multi-level data processing, it can significantly improve the accuracy and efficiency of detection, providing a new solution for infrared detection.

[0004] In view of this, one aspect of the present invention proposes an intelligent analysis method for medical data based on infrared images, including: Obtaining first infrared image data collected by a first infrared detection terminal; Determining a detection target and a detection scheme of a second infrared detection terminal according to the first infrared image data and a preset infrared detection model; Controlling the second infrared detection terminal to perform detection according to the detection target and the detection scheme to obtain second infrared image data; Analyzing the second infrared image by using a preset infrared image analysis model to obtain an infrared image analysis result; Determining an auxiliary imaging detection scheme according to the detection target and the physical data of the object to be detected; Detecting the object to be detected according to the auxiliary imaging detection scheme to obtain auxiliary imaging data; Analyzing the auxiliary imaging data by using a preset auxiliary imaging analysis model to obtain an auxiliary imaging analysis result; Obtaining a comprehensive analysis result according to the infrared image analysis result and the auxiliary imaging analysis result.

[0005] Optionally, the step of determining a detection target and a detection scheme of a second infrared detection terminal according to the first infrared image data and a preset infrared detection model includes: Feature extraction is performed on the first infrared image data, including: extracting the first spatial feature of the temperature distribution; extracting the first temporal feature of the temperature change; calculating the first temperature gradient feature of the key area; The first spatial feature, the first temporal feature, and the first temperature gradient feature are input into a preset infrared detection model for analysis to obtain a first analysis result; wherein, the infrared detection model includes: a suspected anomaly feature recognition sub-model; a risk level assessment sub-model; a detection recommendation generation sub-model; Based on the first analysis result, the detection target is determined, including: identifying the abnormal temperature distribution area; determining the potential lesion location; dividing the detection priority level; A detection plan is generated according to the detection target, including: determining the detection scope and location; selecting the corresponding detection parameter settings; formulating the detection timing arrangement; The detection plan is optimized and adjusted, including: making personalized adjustments in combination with the physical condition of the object to be examined; conducting a feasibility assessment considering the technical characteristics of the detection equipment; adjusting the detection order according to the urgency; A detection task instruction is generated, including: converting the detection plan into an instruction set executable by the equipment; configuring the detection parameters and control parameters; generating a detection process description document.

[0006] Optionally, the step of controlling the second infrared detection terminal to perform detection according to the detection target and the detection plan to obtain the second infrared image data includes: Performing self-check and calibration of the second infrared detection terminal, including: checking the working status and environmental parameters of the second infrared detection terminal; performing temperature calibration on the infrared detector of the second infrared detection terminal; verifying the resolution and sensitivity of the image acquisition system of the second infrared detection terminal; Configuring the detection parameters according to the detection target and the detection plan, including: setting the working mode of the infrared detector; adjusting the resolution and frame rate of image acquisition; configuring the temperature measurement range and accuracy; Controlling the second infrared detection terminal to perform detection preparation according to the configured detection parameters, including: adjusting the spatial position of the second infrared detection terminal; setting the optimal imaging distance; optimizing the detection angle; Performing multi-mode scanning on the target area, including: performing conventional infrared scanning; performing high-resolution local scanning; collecting dynamic temperature change data; Real-time monitoring of the detection process, including: tracking the real-time changes of the detection parameters; evaluating the image quality; timely adjusting the detection strategy; Processing the collected infrared image data to obtain the second infrared image data, including: performing image enhancement and noise reduction; performing geometric correction and registration; generating a temperature distribution map.

[0007] Optionally, the step of analyzing the second infrared image using a preset infrared image analysis model to obtain an infrared image analysis result includes: Loading a preset infrared image analysis model, including: a hot spot recognition model; an abnormal feature extraction model; an abnormal pattern classification model; Performing feature extraction on the second infrared image to obtain a second feature set, including: extracting second spatial features of the temperature distribution; extracting second morphological features of the hot spot region; extracting second boundary features of temperature anomalies; calculating regional temperature statistical features; Inputting the second feature set into the infrared image analysis model to perform lesion area recognition and analysis, including: locating the abnormal temperature area; analyzing the morphological features of the lesion area; calculating the area and depth of the lesion area; evaluating the activity level of the lesion area; The infrared image analysis model performs abnormal pattern matching according to the second feature set, including: matching with a standard abnormal pattern library; calculating the matching probabilities of various abnormal types; generating preliminary analysis suggestions; Performing multi-dimensional abnormal evaluation, including: evaluating the development stage of the lesion; analyzing the spread trend of the lesion; predicting the risk of disease development; Generating an analysis report as the infrared image analysis result, including: summarizing the quantitative analysis results of the abnormal area; generating a three-dimensional schematic diagram of the lesion area; outputting analysis suggestions and risk warnings.

[0008] Optionally, the step of determining an auxiliary imaging detection scheme according to the detection target and the physical data of the object to be examined includes: Collecting and sorting out the physical data of the object to be examined, including: extracting basic physiological indicators; collecting past medical history information; recording current symptom descriptions; sorting out relevant examination results; Performing classification analysis on the detection target, including: determining the type of the target tissue; analyzing the depth position of the lesion; evaluating the size range of the lesion; judging the activity level of the lesion; Establishing a detection adaptability evaluation model, including: evaluating the applicability of various imaging techniques; analyzing detection risks and detection limiting conditions; calculating the detection benefit ratio; predicting the detection difficulty; Generating candidate detection schemes according to the physical data, the analysis results of the detection target, and the detection adaptability evaluation model, including: listing optional imaging detection methods; configuring detection parameter suggestions; designing detection timing arrangements; estimating detection costs; Performing scheme optimization and selection, including: evaluating the feasibility of each scheme; comparing the expected detection effects; weighing the detection cost-benefit; evaluating the acceptance of the object to be examined; Output the final detection plan as an auxiliary image detection plan, including: determining the main detection method; formulating alternative detection plans; generating detection precautions; and compiling a detection process description.

[0009] Optionally, the step of analyzing the auxiliary image data by using a preset auxiliary image analysis model to obtain an auxiliary image analysis result includes: Loading a preset auxiliary image analysis model, including: an image preprocessing sub-model; an organ segmentation sub-model; an abnormality recognition sub-model; and an abnormality analysis sub-model; Performing image preprocessing on the auxiliary image data by using the image preprocessing sub-model to obtain first auxiliary image data, including: removing image noise; enhancing image contrast; correcting image distortion; and unifying image specifications; Performing organ and tissue segmentation on the first auxiliary image data by using the organ segmentation sub-model to obtain second auxiliary image data, including: identifying anatomical structures; segmenting target regions; extracting boundary information; and marking key parts; Performing abnormal feature analysis on the second auxiliary image data by using the abnormality recognition sub-model to obtain third auxiliary image data, including: detecting abnormal regions; extracting lesion features; measuring lesion sizes; and evaluating lesion properties; Performing multi-dimensional analysis on the third auxiliary image data by using the abnormality analysis sub-model, including: analyzing lesion distribution; evaluating lesion progression; comparing historical data; and generating analysis suggestions; Outputting an auxiliary image analysis result, including: sorting out quantitative indicators; generating image annotations; summarizing analysis conclusions; and providing reference suggestions.

[0010] Optionally, the step of obtaining a comprehensive analysis result according to the infrared image analysis result and the auxiliary image analysis result includes: Performing data alignment and fusion on the infrared image analysis result and the auxiliary image analysis result, including: unifying data formats and coordinate systems; registering infrared and auxiliary image data; establishing corresponding relationships of feature points; and generating a fusion data set; Constructing a multi-modal analysis model, including: loading a feature correlation model; configuring an analysis rule base; initializing a decision support system; and setting weight parameters; Performing feature correlation analysis on the fusion data set, including: extracting complementary feature information; analyzing the degree of feature correlation; identifying feature conflict points; and synthesizing key features; Performing comprehensive abnormality evaluation according to the multi-modal analysis model and the key features, including: integrating multi-source abnormality information; evaluating the degree of evidence support; analyzing the reliability of abnormality evaluation; and generating a comprehensive evaluation opinion; Combining the comprehensive evaluation opinion, performing risk assessment and prediction, including: analyzing the abnormal development trend and formulating a risk warning; Generate a health analysis report, including: summarizing key abnormal findings; generating multimodal images; providing health analysis suggestions; formulating a follow-up plan.

[0011] Optionally, the method for constructing the infrared detection model includes: Construct a quantum deep learning framework, including: designing the structure of the quantum convolutional layer; constructing the quantum pooling operation; defining the quantum activation function; configuring the quantum fully connected layer; Design a hybrid quantum-classical architecture, including: constructing a quantum feature extraction module; designing a classical neural network module; defining the interface protocol between modules; implementing a data conversion mechanism; Prepare a training data set, including: collecting infrared detection operation data, infrared detection device working data, infrared detection object data, and infrared detection working environment data; annotating abnormal feature information; dividing data subsets; generating quantum state representations; Execute the model training process, including: initializing quantum parameters; performing quantum state evolution; calculating the loss function; updating model parameters; Implement model optimization to obtain the final infrared detection model, including: analyzing the impact of quantum noise; optimizing the depth of the quantum circuit; adjusting the hyperparameter configuration; performing model pruning; Evaluate the model performance, including: testing the detection accuracy; evaluating the consumption of computing resources; analyzing the model stability; verifying the generalization ability.

[0012] Optionally, the method for constructing the infrared image analysis model includes: Construct a federated learning network architecture, including: deploying a central coordination server; configuring local training nodes; establishing a secure communication channel; designing a data synchronization mechanism; Initialize the local analysis model, including: defining the model structure; configuring privacy protection parameters; setting training hyperparameters; initializing the local data set; Execute federated model training, including: performing local model training using the collected infrared image sample data; encrypting model parameters; uploading parameter updates; aggregating the global model; Implement differential privacy protection, including: calculating the privacy budget; adding noise perturbation; verifying the degree of privacy protection; adjusting the protection parameters; Optimize the model performance to obtain the infrared image analysis model, including: evaluating the model convergence; analyzing the parameter distribution; adjusting the aggregation strategy; optimizing the communication efficiency; Deploy and apply the model, including: verifying the model effect; deploying local services; monitoring the running status; updating the model version.

[0013] Another aspect of the present invention provides an intelligent medical data analysis system based on infrared images, which is used to execute an intelligent medical data analysis method based on infrared images, and is characterized by including: a server, a first infrared detection terminal, and a second infrared detection terminal; The server is configured to: Obtain the first infrared image data collected by the first infrared detection terminal; Determine the detection target and detection plan of the second infrared detection terminal according to the first infrared image data and a preset infrared detection model; Control the second infrared detection terminal to perform detection according to the detection target and the detection plan to obtain the second infrared image data; Analyze the second infrared image by using a preset infrared image analysis model to obtain an infrared image analysis result; Determine an auxiliary imaging detection plan according to the detection target and the physical data of the object to be examined; Detect the object to be examined according to the auxiliary imaging detection plan to obtain auxiliary imaging data; Analyze the auxiliary imaging data by using a preset auxiliary imaging analysis model to obtain an auxiliary imaging analysis result; Obtain a comprehensive analysis result according to the infrared image analysis result and the auxiliary imaging analysis result.

[0014] Adopting the technical solution of the present invention, an intelligent analysis method for medical data based on infrared images includes: acquiring first infrared image data collected by a first infrared detection terminal; determining the detection target and detection plan of a second infrared detection terminal according to the first infrared image data and a preset infrared detection model; controlling the second infrared detection terminal to perform detection according to the detection target and the detection plan to obtain second infrared image data; analyzing the second infrared image by using a preset infrared image analysis model to obtain an infrared image analysis result; determining an auxiliary imaging detection plan according to the detection target and the physical data of the object to be examined; performing detection on the object to be examined according to the auxiliary imaging detection plan to obtain auxiliary imaging data; analyzing the auxiliary imaging data by using a preset auxiliary imaging analysis model to obtain an auxiliary imaging analysis result; and obtaining a comprehensive analysis result according to the infrared image analysis result and the auxiliary imaging analysis result. By combining the detection targets and plans of the first infrared detection terminal and the second infrared detection terminal, accurate detection of the object to be examined can be achieved. This method ensures the pertinence and effectiveness of the detection process; by using the infrared image analysis model and the auxiliary imaging analysis model to analyze different types of data respectively, a more comprehensive interpretation of medical data can be provided. This multi-level analysis method helps to improve the accuracy of diagnosis; determining the auxiliary imaging detection method according to the physical data of the object to be examined makes the detection process more personalized and flexible, and can adapt to the specific needs of different patients; by integrating the infrared image analysis result and the auxiliary imaging analysis result, a more comprehensive and accurate comprehensive analysis result can be obtained; this method can improve the detection efficiency, reduce the waiting time of the object to be examined, and improve the service quality through an intelligent detection and analysis process. The solution of this embodiment can significantly improve the accuracy and efficiency of detection through an intelligent analysis method and multi-level data processing, providing a new solution for infrared detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of an intelligent analysis method for medical data based on infrared images provided by an embodiment of the present invention; Figure 2 is a schematic block diagram of an intelligent analysis system for medical data based on infrared images provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0017] In the following description, numerous specific details are set forth to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways than those specifically described herein, and thus, the scope of protection of the present invention is not limited by the specific embodiments disclosed below.

[0018] The terms "first", "second", etc. in the specification, claims and drawings of this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0019] Reference herein to "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment can be included in at least one embodiment of this application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0020] The following refers to Figures 1 to 2 to describe a method and system for intelligent analysis of medical data based on infrared images provided according to some embodiments of the present invention.

[0021] As Figure 1 shown, an embodiment of the present invention provides a method for intelligent analysis of medical data based on infrared images, including: Obtaining first infrared image data collected by a first infrared detection terminal (such as wearable clothes, vests, patches, etc.); It can be understood that the first infrared detection terminal can be wearable clothes, vests, patches, etc. Determine the layout scheme of infrared sensors according to the preset detection parts, and integrate multiple infrared sensors into the first infrared detection terminal; collect the ambient temperature data in the wearing state of the device, and establish a temperature reference value; calibrate the infrared sensor array, including: collecting the infrared image data of the standard temperature source, and establishing a temperature correction model according to the corresponding relationship between the actual temperature value of the standard temperature source and the infrared image data; collect the infrared image data of the target detection area, including: continuously collecting multiple frames of infrared images, performing motion artifact correction on the collected infrared images, and calibrating the infrared image data according to the temperature correction model; preprocess the calibrated infrared image data, including: removing image noise and outliers, enhancing image contrast, and extracting the temperature distribution characteristics of the target area; perform time series analysis on the preprocessed infrared image data, including: calculating the temperature change trend of the target area, identifying abnormal temperature change patterns, and generating a temperature distribution heat map. In this step, through the collaborative work of multiple sensors, the acquisition accuracy of temperature data is improved; the environmental temperature compensation and calibration process ensure the reliability of the data; the motion artifact correction improves the data quality during dynamic acquisition; the first infrared detection terminal allows long-term continuous monitoring, the time series analysis can timely detect abnormal changes, and the heat map intuitively displays the temperature distribution; automatically complete data preprocessing and feature extraction, with the ability to identify abnormal patterns, which can provide a decision-making basis for subsequent precise detection; the wearable form reduces the restriction on the activities of the detected person, with a high degree of automation, simple operation, and real-time acquisition of detection data.

[0022] Determine the detection target and detection scheme of the second infrared detection terminal according to the first infrared image data and the preset infrared detection model; It can be understood that the detection target may include the detected parts, areas, detection indicators, detection standards, etc.

[0023] Control the second infrared detection terminal (a large infrared detection device capable of precise detection) to perform detection according to the detection target and the detection scheme, and obtain the second infrared image data; Analyze the second infrared image using the preset infrared image analysis model to obtain the infrared image analysis result; Determine the auxiliary imaging detection scheme according to the detection target and the physical data of the object to be detected; Detect the object to be detected according to the auxiliary imaging detection scheme to obtain the auxiliary imaging data; It is understandable that according to the selected auxiliary image detection method, the corresponding detection equipment is prepared (for example, if magnetic resonance imaging is selected, a magnetic resonance instrument needs to be prepared; if X-ray is selected, an X-ray machine needs to be prepared); according to the physical data and detection target of the object to be examined, the relevant parameters of the detection equipment are set, which includes adjusting the sensitivity, resolution, acquisition frequency, etc. of the equipment to ensure that clear image data can be obtained; the object to be examined is positioned within the effective detection range of the detection equipment to ensure that the target area of the object to be examined can be accurately captured by the equipment; the detection equipment is started, and the image of the object to be examined is collected according to the preset detection plan. During this process, the equipment will record the image data of the object to be examined in real time; the collected auxiliary image data is stored in the data processing system, and the data is preliminarily processed, including denoising, enhancement, and format conversion, etc., for subsequent analysis; the processed auxiliary image data is transmitted to the central database or analysis platform for further analysis and interpretation. In this step, by reasonably setting the detection parameters and equipment, high-quality auxiliary image data can be obtained, providing a reliable basis for subsequent analysis; during the detection process, the state of the object to be examined can be monitored in real time to ensure the timeliness and accuracy of the data; through systematic data storage and processing, the collected image data is ensured to be complete and easy for subsequent analysis; the obtained auxiliary image data can provide important reference information for doctors.

[0024] The auxiliary image data is analyzed using a preset auxiliary image analysis model to obtain an auxiliary image analysis result; According to the infrared image analysis result and the auxiliary image analysis result, a comprehensive analysis result is obtained.

[0025] In the embodiment of the present invention, by combining the detection targets and schemes of the first infrared detection terminal and the second infrared detection terminal, accurate detection of the object to be examined can be achieved. This method ensures the pertinence and effectiveness of the detection process; using the infrared image analysis model and the auxiliary image analysis model to analyze different types of data respectively can provide a more comprehensive interpretation of medical data. This multi-level analysis method helps to improve the accuracy of diagnosis; determining the auxiliary image detection method according to the physical data of the object to be examined makes the detection process more personalized and flexible, and can adapt to the specific needs of different patients; by integrating the infrared image analysis result and the auxiliary image analysis result, a more comprehensive and accurate comprehensive analysis result can be obtained; this method can improve the detection efficiency through an intelligent detection and analysis process, reduce the waiting time of the object to be examined, and improve the service quality. The solution of this embodiment can significantly improve the accuracy and efficiency of detection through an intelligent analysis method and multi-level data processing, providing a new solution for infrared detection.

[0026] In some possible embodiments of the present invention, the step of determining the detection target and detection scheme of the second infrared detection terminal according to the first infrared image data and a preset infrared detection model includes: Extract features from the first infrared image data, including: extracting the first spatial feature of the temperature distribution; extracting the first temporal feature of the temperature change; calculating the first temperature gradient feature of the key area; Input the first spatial feature, the first temporal feature, and the first temperature gradient feature into a preset infrared detection model for analysis to obtain a first analysis result; wherein, the infrared detection model includes: a suspected abnormal feature recognition sub-model; a risk level assessment sub-model; a detection suggestion generation sub-model; It can be understood that inputting the first spatial feature, the first temporal feature, and the first temperature gradient feature into the suspected abnormal feature recognition sub-model can obtain a suspected abnormal feature recognition result; inputting the suspected abnormal feature recognition result into the risk level assessment sub-model can obtain a risk level assessment result of the suspected abnormal feature; according to the risk level assessment result and the detection suggestion generation sub-model, a detection suggestion for further detection can be obtained; integrating the suspected abnormal feature recognition result, the risk level assessment result, and the detection suggestion into a first analysis result.

[0027] Determine the detection target based on the first analysis result, including: identifying the abnormal temperature distribution area; determining the potential lesion location; dividing the detection priority level; It can be understood that in this step, the abnormal temperature distribution area can be identified and the potential lesion location can be determined according to the abnormal feature recognition result, and the detection priority level can be divided according to the risk level assessment result and the detection suggestion, so as to obtain the detection target.

[0028] Generate a detection scheme according to the detection target, including: determining the detection range and part; selecting the corresponding detection parameter settings; formulating the detection time sequence arrangement; In this step, determine the detection range, part, and select the corresponding detection parameter settings according to the abnormal temperature distribution area and the potential lesion location, and formulate the detection time sequence arrangement according to the detection priority level, so as to obtain the detection scheme.

[0029] Optimize and adjust the detection scheme, including: making personalized adjustments in combination with the physical condition of the object to be detected; conducting a feasibility assessment considering the technical characteristics of the detection equipment; adjusting the detection order according to the urgency; Generate a detection task instruction, including: converting the detection scheme into an instruction set executable by the device; configuring the detection parameters and control parameters; generating a detection process description document.

[0030] The solution of this embodiment accurately locates the areas that need to be focused on through multi-dimensional feature analysis; the personalized detection solution improves the pertinence of detection; the systematic parameter configuration ensures the detection quality; the model automatically analyzes and evaluates the risk level, intelligently generates the optimal detection solution, and automatically adapts to different detection scenarios; the priority assignment improves the detection efficiency, reasonably arranges the detection order, and avoids unnecessary repeated detections; the standardized detection process, clear execution instructions, and traceable detection records.

[0031] In some possible embodiments of the present invention, the step of controlling the second infrared detection terminal to perform detection according to the detection target and the detection solution to obtain the second infrared image data includes: Perform self-check and calibration of the second infrared detection terminal, including: checking the working status and environmental parameters of the second infrared detection terminal; performing temperature calibration of the infrared detector of the second infrared detection terminal; verifying the resolution and sensitivity of the image acquisition system of the second infrared detection terminal; Configure detection parameters according to the detection target and the detection solution, including: setting the working mode of the infrared detector; adjusting the resolution and frame rate of image acquisition; configuring the temperature measurement range and accuracy; Control the second infrared detection terminal to perform detection preparation according to the configured detection parameters, including: adjusting the spatial position of the second infrared detection terminal; setting the optimal imaging distance; optimizing the detection angle; Perform multi-mode scanning on the target area, including: performing conventional infrared scanning; performing high-resolution local scanning; collecting dynamic temperature change data; Monitor the detection process in real time, including: tracking the real-time changes of detection parameters; evaluating the image quality; adjusting the detection strategy in a timely manner; Process the collected infrared image data to obtain the second infrared image data, including: performing image enhancement and noise reduction; performing geometric correction and registration; generating a temperature distribution map.

[0032] The solution of this embodiment ensures the basic detection performance through device self-check and calibration, provides more comprehensive detection data through multi-mode scanning, and ensures data reliability through real-time quality monitoring; high-resolution local scanning provides more detailed lesion information, dynamic temperature monitoring reflects the tissue function state, and the accurate temperature distribution map helps with diagnostic analysis; the standardized detection process, real-time parameter monitoring and adjustment, and the detection process are traceable; multi-dimensional data collection, systematic data processing, and standardized data storage format.

[0033] In some possible embodiments of the present invention, the step of analyzing the second infrared image using a preset infrared image analysis model to obtain the infrared image analysis result includes: Load the pre-set infrared image analysis models, including: a hot spot recognition model; an abnormal feature extraction model; an abnormal pattern classification model; It can be understood that the infrared image analysis models include, but are not limited to, a hot spot recognition model, an abnormal feature extraction model, and an abnormal pattern classification model. Among them, the hot spot recognition model is used to identify hot spot areas in the image, usually referring to areas with higher temperature or abnormal activity; in medical applications, hot spots may indicate potential lesions or inflammation areas; by analyzing infrared images, the hot spot recognition model can quickly locate these key areas and provide important clues for subsequent diagnosis. The main function of the abnormal feature extraction model is to extract features significantly different from the normal state from infrared images. These features may include temperature changes, abnormal shapes, or other health-related indicators. By analyzing the image data, this model helps identify possible abnormal conditions and provides basic data for subsequent analysis and decision-making. The abnormal pattern classification model is used to classify the extracted abnormal features and determine whether they belong to a specific abnormal pattern. Through training, the model can identify different types of abnormalities (such as tumors, inflammation, etc.) and classify them. This process usually involves machine learning algorithms, which can improve the accuracy and efficiency of classification. The combined use of these models makes infrared image analysis more systematic and intelligent, improving the accuracy and efficiency of medical diagnosis.

[0034] Extract features from the second infrared image to obtain a second feature set, including: extracting the second spatial feature of the temperature distribution; extracting the second morphological feature of the hot spot area; extracting the second boundary feature of the temperature anomaly; calculating the regional temperature statistical feature; Input the second feature set into the infrared image analysis model to perform lesion area recognition and analysis, including: locating the abnormal temperature area; analyzing the morphological features of the lesion area; calculating the area and depth of the lesion area; evaluating the activity level of the lesion area; In this step, locate the abnormal temperature area through the hot spot recognition model; use the abnormal feature extraction model to analyze the morphological features of the lesion area, calculate the area and depth of the lesion area, and evaluate the activity level of the lesion area.

[0035] The infrared image analysis model performs abnormal pattern matching according to the second feature set, including: matching with a standard abnormal pattern library; calculating the matching probabilities of various abnormal types; generating preliminary analysis suggestions; Perform multi-dimensional abnormal evaluation, including: evaluating the development stage of the lesion; analyzing the spread trend of the lesion; predicting the risk of disease development; Generate an analysis report as the result of the infrared image analysis, including: summarizing the quantitative analysis results of the abnormal areas; generating a three-dimensional schematic diagram of the lesion area; outputting analysis suggestions and risk warnings.

[0036] The solution of this embodiment improves the judgment accuracy through multi-model collaborative analysis, provides an objective evaluation basis through quantitative analysis, and provides a reliable judgment reference through pathological pattern matching; automatically identifies and locates abnormal areas, intelligently generates recognition results, and provides early warnings through predictive analysis; extracts multi-dimensional features, comprehensively evaluates the condition, and outputs a complete analysis report.

[0037] In some possible embodiments of the present invention, the step of determining an auxiliary imaging detection scheme according to the detection target and the physical data of the object to be examined includes: Collect and sort out the physical data of the object to be examined, including: extracting basic physiological indicators; collecting past medical history information; recording current symptom descriptions; sorting out relevant examination results; Conduct a classification analysis of the detection target, including: determining the type of the target tissue; analyzing the depth position of the lesion; evaluating the size range of the lesion; judging the activity degree of the lesion; It can be understood that the detection targets include, but are not limited to, determining the type of the target tissue, analyzing the depth position of the lesion, evaluating the size range of the lesion, and judging the activity degree of the lesion. Among them, determining the type of the target tissue involves identifying different tissue types that appear, such as normal tissue, tumor tissue, inflammatory tissue, etc.; analyzing the depth position of the lesion aims to determine the depth position of the lesion in the tissue, that is, whether the lesion is superficial or deep, which is crucial for evaluating the severity of the lesion and formulating a treatment plan. For example, some lesions may be limited to the surface layer, while other lesions may have invaded deeper tissue layers; evaluating the size range of the lesion refers to measuring the diameter or area of the lesion area, and this information helps to understand the extent of the lesion's expansion, thereby judging its possible impact and the urgency of treatment; judging the activity degree of the lesion involves evaluating the biological activity of the lesion, such as whether it is growing, whether there is an inflammatory reaction, etc. By analyzing the temperature change, morphological characteristics, etc. of the lesion, it can be judged whether the lesion is in an active state.

[0038] Establish a detection adaptability evaluation model, including: evaluating the applicability of various imaging techniques; analyzing detection risks and detection limiting conditions; calculating the detection benefit ratio; predicting the detection difficulty; It is understandable that the process of establishing a detection adaptability evaluation model aims to comprehensively evaluate the effectiveness and applicability of different imaging techniques in specific detection tasks. Specifically, evaluating the applicability of multiple imaging techniques involves comparing and evaluating available imaging techniques (such as infrared thermography, X-ray, CT, MRI, etc.) to determine which technique is most suitable for specific detection objectives and conditions. The evaluation criteria may include factors such as image quality, resolution, detection speed, cost, and patient comfort. Analyzing detection risks and detection limiting conditions means that during the detection process, there may be certain risks and limiting conditions, such as radiation exposure, physiological effects on patients, and equipment availability. Analyzing these risks and limiting conditions helps ensure the safety and effectiveness of the detection process and provides necessary risk assessment information. Calculating the detection benefit ratio: The detection benefit ratio refers to the ratio between the benefits obtained through detection (such as early disease detection, improved treatment effects, etc.) and the costs required for detection (such as time, money, resources, etc.). Calculating this ratio helps evaluate the economy and practicality of the detection and helps decision-makers select the most cost-effective detection plan. Predicting the detection difficulty involves evaluating the complexity and challenges of the detection process, including technical difficulty, operation requirements, and complexity of data analysis. Predicting the detection difficulty can help the medical team prepare well and ensure that possible problems can be effectively addressed during the implementation of the detection. By establishing such a detection adaptability evaluation model, it is possible to more scientifically select and implement imaging detection techniques, thereby improving the accuracy and efficiency of diagnosis, optimizing resource allocation, and ensuring patient safety.

[0039] Generate a candidate detection plan based on the body data, the analysis result of the detection target, and the detection adaptability evaluation model, including: listing optional imaging detection methods; configuring detection parameter suggestions; designing a detection time sequence arrangement; estimating the detection cost; It is understandable that listing optional imaging detection methods means listing suitable imaging detection techniques according to the patient's body data and the characteristics of the detection target. For example, possible options may include infrared thermography, X-ray, CT scan, MRI, etc. Each technique has its unique advantages and disadvantages and is suitable for different detection needs. Configuring detection parameter suggestions means providing specific detection parameter suggestions for each imaging detection method. These parameters may include imaging resolution, exposure time, scanning speed, use of contrast agent, etc. to ensure the accuracy and effectiveness of the detection. Designing a detection time sequence arrangement means formulating a reasonable detection time sequence arrangement, considering the patient's time arrangement, the availability of detection equipment, and the analysis time of the detection results. A reasonable time sequence arrangement can improve the detection efficiency and reduce the patient's waiting time. Estimating the detection cost means estimating the cost of each candidate detection plan, including equipment usage fees, material costs (such as contrast agent), labor costs, etc. This estimation helps evaluate the economy of different plans and ensures the selection of the best plan within the budget.

[0040] Conduct optimization and selection of implementation plans, including: evaluating the feasibility of each plan; comparing the expected detection effects; weighing the cost-benefit of detection; evaluating the acceptance of the objects to be detected; It can be understood that evaluating the feasibility of each plan involves analyzing the technical, economic, and operational feasibility of each candidate plan, and it is necessary to consider whether the plan can be effectively implemented under existing conditions, including technical requirements, resource availability, and time constraints, etc. Comparing the expected detection effects requires comparing the expected detection effects of each plan, which includes evaluating the performance of each plan in terms of accuracy, sensitivity, specificity, etc., to determine which plan can provide the best detection results. Weighing the cost-benefit of detection: involves weighing the cost of each plan against its expected benefits, and it is necessary to calculate the total cost of each plan, including equipment, materials, labor, etc. costs, and compare it with the potential benefits it brings to ensure that the selected plan is economically reasonable. Evaluating the acceptance of the objects to be detected: it is necessary to consider the acceptance of the objects to be detected for each plan, which includes the comfort of patients with the detection method, the understanding and trust of the detection process, etc. The acceptance of patients directly affects the smooth progress of the detection and the reliability of the results. Through this step, a scientific and reasonable choice can be made among multiple plans, thereby improving the efficiency and accuracy of the detection and ensuring the safety and satisfaction of patients.

[0041] Output the final detection plan as an auxiliary imaging detection plan, including: determining the main detection method; formulating alternative detection plans; generating detection precautions; compiling a detection process description.

[0042] The solution of this embodiment customizes the detection plan according to the specific situation of the object to be detected, comprehensively considers multiple influencing factors, and provides flexible detection options; based on objective evaluation of data, multi-dimensional plan comparison, and systematic decision-making process; comprehensive risk assessment, clear limit prompts, and complete preventive measures; reasonably allocate detection resources, optimize detection costs, and improve detection efficiency.

[0043] In some possible implementation manners of the present invention, the step of analyzing the auxiliary image data by using a preset auxiliary image analysis model to obtain an auxiliary image analysis result includes: Load a preset auxiliary image analysis model, including: an image preprocessing sub-model; an organ segmentation sub-model; an abnormality recognition sub-model; an abnormality analysis sub-model; It is understandable that the image preprocessing sub-model is responsible for preprocessing the input medical images to improve the accuracy of subsequent analysis. The preprocessing steps may include noise removal, contrast enhancement, standardization of image size, etc. These operations help to eliminate interference factors in the images and make subsequent analysis more effective. The main function of the organ segmentation sub-model is to identify and segment specific organs or tissue structures from medical images. Through precise segmentation, the morphology and location of the target organ can be observed more clearly, providing important information for subsequent diagnosis and treatment. The abnormal recognition sub-model is used to detect possible abnormal or diseased areas in the images. By analyzing image features, this model can identify areas different from normal tissues, thus prompting doctors to pay attention to potential health problems. Abnormal analysis sub-model: After identifying the abnormalities, this sub-model further analyzes the characteristics of these abnormalities, including size, shape, location, etc. This analysis helps to evaluate the nature and severity of the abnormalities, so as to formulate corresponding plans. Through the collaborative work of these sub-models, the auxiliary image analysis model can effectively improve the analysis efficiency and accuracy of medical images, helping doctors to make more accurate diagnosis and treatment decisions.

[0044] Perform image preprocessing on the auxiliary image data using the image preprocessing sub-model to obtain the first auxiliary image data, including: removing image noise; enhancing image contrast; correcting image distortion; unifying image specifications; Perform organ and tissue segmentation using the organ segmentation sub-model and the first auxiliary image data to obtain the second auxiliary image data, including: identifying anatomical structures; segmenting target regions; extracting boundary information; marking key parts; Perform abnormal feature analysis using the abnormal recognition sub-model and the second auxiliary image data to obtain the third auxiliary image data, including: detecting abnormal regions; extracting lesion features; measuring lesion size; evaluating lesion nature; Perform multi-dimensional analysis using the abnormal analysis sub-model and the third auxiliary image data, including: analyzing lesion distribution; evaluating lesion progression; comparing historical data; generating analysis suggestions; Output the auxiliary image analysis results, including: sorting out quantitative indicators; generating image annotations; summarizing analysis conclusions; providing reference suggestions.

[0045] The solution of this embodiment can quickly process and analyze a large amount of auxiliary image data using a preset analysis model, improving the analysis efficiency; through in-depth analysis of the auxiliary image data, potential health problems can be accurately identified, helping doctors to make more precise analysis; reducing the need for manual intervention, reducing the possibility of human errors, and improving the consistency and reliability of the analysis results; being able to provide doctors with real-time analysis results to support quick decision-making and timely intervention.

[0046] In some possible embodiments of the present invention, the step of obtaining a comprehensive analysis result according to the infrared image analysis result and the auxiliary image analysis result includes: Performing data alignment and fusion on the infrared image analysis result and the auxiliary image analysis result, including: unifying the data format and coordinate system; registering the infrared and auxiliary image data; establishing a corresponding relationship of feature points; generating a fused data set; Constructing a multi-modal analysis model, including: loading a feature correlation model; configuring an analysis rule library; initializing a decision support system; setting weight parameters; It can be understood that loading the feature correlation model involves integrating the feature correlation model into a multi-modal analysis framework. The feature correlation model is used to identify and quantify the relationships and interactions between different data sources (such as image data, clinical data, etc.), which helps to understand the importance of each feature in the overall analysis. Configuring the analysis rule library: Here, an analysis rule library needs to be established, which contains various rules and standards for data analysis. These rules can be based on domain knowledge, historical data analysis results, or expert experience, and are designed to guide the model on how to process and interpret different types of data. Initializing the decision support system means setting up a system that can provide suggestions and decision support based on the analysis results. This system usually integrates the analysis results of multi-modal data and provides actionable suggestions for users (such as doctors or decision-makers) to help them make more informed decisions. Setting weight parameters: In multi-modal analysis, the importance of different data sources and features may vary, so weight parameters need to be set for each feature. These weight parameters can be optimized through machine learning algorithms to ensure that the model can fully consider the influence of each feature during analysis, thereby improving the accuracy of prediction. Through this step, the constructed multi-modal analysis model can effectively integrate and analyze data from different sources, provide more comprehensive insights and support, and help users make better decisions.

[0047] Performing feature correlation analysis on the fused data set, including: extracting complementary feature information; analyzing the degree of feature correlation; identifying feature conflict points; synthesizing key features; It is understandable that performing feature correlation analysis on the fused dataset aims to extract valuable information from multiple data sources to improve the performance and accuracy of the model. Specifically, complementary feature information is extracted: in the fused dataset, different features may provide complementary information. By analyzing these features, it is possible to identify which features provide unique perspectives in different data sources, thereby enhancing the expressive power of the overall dataset. Analyze the degree of feature correlation: This involves evaluating the correlation between different features. By calculating the correlation coefficient between features or using other statistical methods, it is possible to determine which features are correlated with each other and which features are independent; this helps to understand the role of features in the dataset and provides a basis for subsequent feature selection. Identify feature conflict points: In the fused dataset, certain features may conflict, that is, the information they provide is contradictory; identifying these conflict points is crucial because they may affect the decision-making process of the model; by analyzing the relationship between features, these conflicts can be identified and corresponding processing can be carried out. Synthesize key features: Through comprehensive analysis of features, key features can be synthesized; these features are usually the ones that best represent the overall information of the dataset and can effectively improve the prediction ability of the model; the process of synthesizing key features may involve methods such as feature selection, dimensionality reduction techniques, or feature engineering. Through this step, feature correlation analysis can help better understand the structure of the dataset, optimize the input features of the model, and thus improve the accuracy of analysis and prediction.

[0048] Perform comprehensive anomaly assessment based on the multi-modal analysis model and the key features, including: integrating multi-source anomaly information; evaluating the degree of evidence support; analyzing the reliability of anomaly assessment; generating a comprehensive assessment opinion; Combined with the comprehensive assessment opinion, perform risk assessment and prediction, including: analyzing the trend of anomaly development and formulating risk warnings; Generate a health analysis report, including: summarizing key anomaly findings; generating multi-modal images; providing health analysis suggestions; formulating a follow-up plan.

[0049] The solution of this embodiment can provide a more comprehensive health assessment by comprehensively analyzing the results of infrared images and auxiliary images, helping doctors better understand the health status of the detected object; considering the analysis results of the two images comprehensively can improve the accuracy of health analysis and reduce the possibility of misjudgment.

[0050] In some possible implementation manners of the present invention, the construction method of the infrared detection model includes: Construct a quantum deep learning framework, including: designing the structure of the quantum convolutional layer; constructing quantum pooling operations; defining quantum activation functions; configuring quantum fully connected layers; In this step, the quantum convolutional layer is a core part of the quantum deep learning framework, aiming to utilize the characteristics of quantum computing to process and extract features of input data; designing the structure of the quantum convolutional layer requires considering how to transform classical convolutional operations into quantum operations, for example, implementing convolutional operations through quantum gates and quantum circuits. Construct the quantum pooling operation: The quantum pooling operation is used to reduce the dimensionality of data while retaining important feature information; in quantum deep learning, the pooling operation can be achieved through the measurement and selective operations of quantum states, aiming to improve computational efficiency and reduce the complexity of subsequent layers. Define the quantum activation function: The activation function is used in deep learning to introduce non-linear characteristics, enabling the model to learn complex patterns; in the quantum deep learning framework, defining the quantum activation function requires considering the characteristics of quantum states, which may involve the design of quantum gates to achieve functions similar to classical activation functions. Configure the quantum fully connected layer: The quantum fully connected layer is a key part connecting the previous layer and the next layer, responsible for integrating the relationships between quantum bits; configuring the quantum fully connected layer requires designing appropriate quantum gates and circuits to ensure that information can be effectively transmitted from one layer to another. Through this step, a powerful quantum deep learning framework can be constructed to utilize the advantages of quantum computing to process complex data tasks, thus promoting the development of machine learning and artificial intelligence.

[0051] Design a hybrid quantum-classical architecture, including: constructing a quantum feature extraction module; designing a classical neural network module; defining the interface protocol between modules; implementing a data conversion mechanism; It is understandable that the main task of the quantum feature extraction module is to extract useful features from the input data, usually implemented through quantum algorithms. The quantum feature extraction module can utilize the parallelism and superposition characteristics of quantum computing to process complex data patterns, thereby improving the efficiency and accuracy of feature extraction. Design the classical neural network module: The classical neural network module is responsible for processing the feature information output by the quantum feature extraction module. This module can adopt traditional neural network architectures, such as fully connected layers, convolutional layers, etc., to further analyze and learn data features. This part usually relies on classical computing resources to perform complex computing tasks. Define the interface protocol between modules: To ensure effective communication between the quantum module and the classical module, clear interface protocols need to be defined. These protocols stipulate how data is transmitted between the two modules, including data formats, transmission methods, and calling methods, etc. This part is the key to implementing the hybrid architecture, ensuring seamless cooperation between different computing platforms. Implement the data conversion mechanism: The data conversion mechanism is used to convert the data format between the quantum feature extraction module and the classical neural network module. This may include converting quantum states into classical data formats, or encoding classical data into quantum states for effective data exchange between the two modules. Through this step, an efficient hybrid quantum-classical architecture can be constructed, making full use of the advantages of quantum computing and the mature technologies of classical computing, thereby promoting the development of machine learning and artificial intelligence.

[0052] Prepare the training dataset, including: collecting infrared detection operation data, infrared detection device working data, infrared detection object data, infrared detection working environment data; annotating abnormal feature information; dividing data subsets; generating quantum state representations; Execute the model training process, including: initializing quantum parameters; performing quantum state evolution; calculating the loss function; updating model parameters; Implement model optimization to obtain the final infrared detection model, including: analyzing the impact of quantum noise; optimizing the depth of the quantum circuit; adjusting the hyperparameter configuration; performing model pruning; Evaluate the model performance, including: testing the detection accuracy; evaluating the computing resource consumption; analyzing the model stability; verifying the generalization ability.

[0053] The solution of this embodiment combines the parallel processing ability of quantum computing and the feature extraction ability of deep learning, which can significantly improve the detection accuracy of the infrared detection model; quantum computing can achieve a faster processing speed than classical computing on specific problems, thereby accelerating the model training and inference processes; through the characteristics of the quantum neural network, it can better process complex data, improve the generalization ability of the model, and reduce the overfitting phenomenon; this model can handle more complex infrared detection tasks, such as target recognition, anomaly detection, etc., expanding the application scope of infrared detection.

[0054] In some possible embodiments of the present invention, the method for constructing the infrared image analysis model includes: Construct a federated learning network architecture, including: deploying a central coordination server; configuring local training nodes; establishing a secure communication channel; designing a data synchronization mechanism; It can be understood that the central coordination server is the core of the federated learning architecture, responsible for coordinating the training processes of each participating node. It does not directly access the original data of the participating nodes, but receives model updates (such as gradients or weights) from each node and aggregates them to generate a global model. This design ensures the privacy and security of the data while improving the training efficiency of the model. Local training nodes refer to the devices or organizations participating in federated learning. They perform model training locally. Each node uses its local data for training and sends the updated trained model to the central server. When configuring these nodes, their computing power, storage resources, and network connections need to be considered to ensure that they can effectively participate in the training process. In federated learning, a secure communication channel is crucial to ensure that data and model updates are not stolen or tampered with during transmission. Encryption technologies (such as homomorphic encryption or secure multi-party computation) can be used to protect the communication content, ensuring that only authorized nodes can access and send information. This measure is essential for protecting user privacy and data security. The data synchronization mechanism is used to ensure data consistency and timeliness between the central server and local training nodes. Since the training processes of each node are independent, an effective synchronization strategy needs to be designed to update the global model in a timely manner after each round of training and distribute the latest model parameters to all participating nodes. This can be achieved through regular model updates and aggregations. Through this step, an efficient and secure federated learning network architecture can be constructed, making full use of decentralized data resources while protecting user privacy and data security.

[0055] Initialize the local analysis model, including: defining the model structure; configuring privacy protection parameters; setting training hyperparameters; initializing the local dataset; It is understandable that defining the model structure involves selecting and designing the architecture of the model, including determining the number of layers in the model, the number of neurons in each layer, activation functions, etc. The design of the model structure directly affects the learning ability and performance of the model. When dealing with sensitive data, privacy protection is of utmost importance; configuring privacy protection parameters may include selecting appropriate encryption methods, setting data access permissions, and defining data processing rules to ensure the security and privacy of user data. Hyperparameters are important parameters that affect the model training process, such as learning rate, batch size, number of training epochs, etc.; setting these hyperparameters reasonably can improve the training efficiency and final performance of the model. The selection of hyperparameters usually needs to be determined through experiments and tuning. The initialization of the local dataset refers to preparing and loading the data used to train the model, which includes data cleaning, preprocessing, and partitioning to ensure the quality and applicability of the data. The quality of the dataset directly affects the training effect and generalization ability of the model. Through this step, the local analysis model can be effectively initialized, laying a solid foundation for subsequent training and evaluation.

[0056] Execute the federated model training, including: performing local model training using the collected infrared image sample data; encrypting the model parameters; uploading the parameter updates; aggregating the global model; Implement differential privacy protection, including: calculating the privacy budget; adding noise perturbation; verifying the degree of privacy protection; adjusting the protection parameters; It is understandable that the privacy budget is a key parameter that measures the strength of differential privacy protection. It determines how much noise can be added in data queries; the smaller the privacy budget, the higher the requirement for privacy protection, but it may reduce the availability and accuracy of the data. In practical applications, the selection of the privacy budget needs to consider the sensitivity of the data and the requirements of the model. To achieve differential privacy, noise usually needs to be added to the query results; common noise addition mechanisms include the Laplace mechanism and the Gaussian mechanism. By introducing noise into the query results, it can be ensured that even if an attacker obtains the query results, they cannot accurately infer the privacy information of individuals. After implementing differential privacy, it is necessary to verify the effectiveness of the privacy protection; this can be achieved by evaluating the performance of the model under different privacy budgets to ensure that the model can still provide useful results while meeting the privacy protection requirements. The verification process may involve comparing the model outputs under different parameter settings to ensure that the degree of privacy protection meets the expectations. According to the verification results, it may be necessary to adjust the privacy protection parameters, such as the privacy budget and the standard deviation of the noise, etc. This process is dynamic and aims to balance the relationship between privacy protection and data availability. By continuously adjusting these parameters, the performance of the model can be optimized while ensuring the effective protection of data privacy. Through this step, differential privacy protection can be effectively implemented to ensure the privacy of users during the data analysis and machine learning processes.

[0057] Optimize the model performance to obtain an infrared image analysis model, including: evaluating model convergence; analyzing parameter distribution; adjusting the aggregation strategy; optimizing communication efficiency; Deploy and apply the model, including: verifying the model effect; deploying local services; monitoring the running status; updating the model version.

[0058] The solution of this embodiment. Federated learning technology allows model training without sharing raw data, protecting user privacy and data security; through the joint training of multi-party data, the model can learn more extensive features and improve its generalization ability in different scenarios; only transmitting model parameters instead of raw data significantly reduces the bandwidth requirements and communication costs of data transmission; distributed training can utilize the computing resources of each participant to accelerate the model training process.

[0059] Please refer to Figure 2 , Another embodiment of the present invention provides an intelligent medical data analysis system based on infrared images for performing an intelligent medical data analysis method based on infrared images, including: a server, a first infrared detection terminal, and a second infrared detection terminal; The server is configured to: Obtain the first infrared image data collected by the first infrared detection terminal; Determine the detection target and detection scheme of the second infrared detection terminal according to the first infrared image data and a preset infrared detection model; Control the second infrared detection terminal to perform detection according to the detection target and the detection scheme to obtain second infrared image data; Analyze the second infrared image using a preset infrared image analysis model to obtain an infrared image analysis result; Determine an auxiliary imaging detection scheme according to the detection target and the physical data of the object to be examined; Perform detection on the object to be examined according to the auxiliary imaging detection scheme to obtain auxiliary imaging data; Analyze the auxiliary imaging data using a preset auxiliary imaging analysis model to obtain an auxiliary imaging analysis result; Obtain a comprehensive analysis result according to the infrared image analysis result and the auxiliary imaging analysis result.

[0060] It should be known that Figure 2 The block diagram of the intelligent medical data analysis system based on infrared images shown is only for illustration, and the number of each module shown does not limit the protection scope of the present invention. The intelligent medical data analysis system based on infrared images provided in this embodiment can be used to execute the solutions of each embodiment of the corresponding intelligent medical data analysis method based on infrared images. For the specific implementation process, please refer to the description of each method embodiment, which will not be elaborated here.

[0061] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0062] In the above embodiments, the descriptions of the respective embodiments have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0063] In the several embodiments provided by this application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the above division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in an electrical or other form.

[0064] The units described as separate components above may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0065] In addition, in each embodiment of this application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0066] When the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned memory includes: various media such as USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), external hard drives, magnetic disks, or optical discs that can store program codes.

[0067] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program. This program can be stored in a computer-readable memory, and the memory can include: flash drives, read-only memories (abbreviation: ROM, English: Read-Only Memory), random access memories (abbreviation: RAM, English: Random Access Memory), magnetic disks, or optical discs, etc.

[0068] The above has introduced the embodiments of this application in detail. Specific examples are used in this article to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to this application.

[0069] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions without departing from the spirit and scope of the present invention, and can make various modifications and alterations, including combinations of the above different functions and implementation steps, including software and hardware implementation manners, all within the protection scope of the present invention.

Claims

1. A medical data intelligent analysis method based on infrared images, characterized in that: include: Acquire first infrared image data collected by a first infrared detection terminal; Determining a detection target and a detection scheme of a second infrared detection terminal according to the first infrared image data and a preset infrared detection model; Controlling the second infrared detection terminal to perform detection according to the detection target and the detection scheme to obtain second infrared image data; Analyzing the second infrared image using a preset infrared image analysis model to obtain an infrared image analysis result; Determining an auxiliary imaging detection scheme according to the detection target and the physical data of the object being detected; Detecting the object to be detected according to the auxiliary image detection scheme to obtain auxiliary image data; Analyzing the auxiliary image data using a preset auxiliary image analysis model to obtain an auxiliary image analysis result; A comprehensive analysis result is obtained based on the infrared image analysis result and the auxiliary image analysis result.

2. The medical data intelligent analysis method based on infrared images according to claim 1 is characterized in that: The step of determining the detection target and detection scheme of the second infrared detection terminal according to the first infrared image data and a preset infrared detection model includes: Extracting features from the first infrared image data includes: extracting a first spatial feature of temperature distribution; extracting a first temporal feature of temperature change; and calculating a first temperature gradient feature of a key area; Inputting the first spatial feature, the first temporal feature and the first temperature gradient feature into a preset infrared detection model for analysis to obtain a first analysis result; wherein the infrared detection model includes: a suspected abnormal feature recognition sub-model; a risk level assessment sub-model; and a detection suggestion generation sub-model; Determining detection targets based on the first analysis results includes: identifying abnormal temperature distribution areas; determining potential lesion locations; and dividing detection priority levels; Generate a detection plan according to the detection target, including: determining the detection range and location; selecting corresponding detection parameter settings; formulating a detection schedule; Optimizing and adjusting the testing plan, including: making personalized adjustments based on the physical condition of the subject; conducting feasibility assessments based on the technical characteristics of the testing equipment; and adjusting the testing sequence based on the degree of urgency; Generate detection task instructions, including: converting the detection plan into an instruction set executable by the device; configuring detection parameters and control parameters; and generating detection process description documents.

3. The medical data intelligent analysis method based on infrared images according to claim 2 is characterized in that: The step of controlling the second infrared detection terminal to perform detection according to the detection target and the detection scheme to obtain second infrared image data includes: Performing self-test and calibration of the second infrared detection terminal, including: checking the working status and environmental parameters of the second infrared detection terminal; performing temperature calibration of the infrared detector of the second infrared detection terminal; and verifying the resolution and sensitivity of the image acquisition system of the second infrared detection terminal; Configuring detection parameters according to the detection target and the detection scheme, including: setting the working mode of the infrared detector; adjusting the resolution and frame rate of image acquisition; configuring the temperature measurement range and accuracy; Controlling the second infrared detection terminal to perform detection preparation according to the configured detection parameters, including: adjusting the spatial position of the second infrared detection terminal; setting the optimal imaging distance; optimizing the detection angle; Perform multi-mode scans on the target area, including: performing conventional infrared scans; performing high-resolution local scans; and collecting dynamic temperature change data; Real-time monitoring of the detection process, including: tracking real-time changes in detection parameters; evaluating image quality; and timely adjusting detection strategies; The collected infrared image data is processed to obtain second infrared image data, including: performing image enhancement and noise reduction; performing geometric correction and registration; and generating a temperature distribution map.

4. The medical data intelligent analysis method based on infrared images according to claim 3 is characterized in that: The step of analyzing the second infrared image using a preset infrared image analysis model to obtain an infrared image analysis result includes: Load the preset infrared image analysis model, including: hotspot recognition model; abnormal feature extraction model; abnormal pattern classification model; Extracting features from the second infrared image to obtain a second feature set includes: extracting a second spatial feature of temperature distribution; extracting a second morphological feature of the hot spot area; extracting a second boundary feature of temperature anomaly; and calculating regional temperature statistical features; Inputting the second feature set into the infrared image analysis model to perform lesion area recognition and analysis, including: locating abnormal temperature areas; analyzing morphological features of lesion areas; calculating the area and depth of lesion areas; and evaluating the activity level of lesion areas; The infrared image analysis model performs abnormal pattern matching according to the second feature set, including: matching with a standard abnormal pattern library; calculating matching probabilities of various abnormal types; and generating preliminary analysis suggestions; Perform multi-dimensional abnormality assessment, including: assessing the development stage of the lesion; analyzing the spread trend of the lesion; predicting the risk of disease progression; Generate an analysis report as the infrared image analysis result, including: summarizing the quantitative analysis results of the abnormal area; generating a three-dimensional schematic diagram of the lesion area; outputting analysis suggestions and risk warnings.

5. The medical data intelligent analysis method based on infrared images according to claim 4 is characterized in that: The step of determining the auxiliary image detection scheme according to the detection target and the body data of the object under inspection includes: Collect and organize the physical data of the subjects, including: extracting basic physiological indicators; collecting past medical history information; recording current symptom descriptions; and organizing relevant examination results; Classifying and analyzing the detection target, including: determining the type of target tissue; analyzing the depth and location of the lesion; evaluating the size range of the lesion; and determining the activity level of the lesion; Establish a test adaptability assessment model, including: assessing the applicability of multiple imaging technologies; analyzing test risks and test restrictions; calculating test benefit ratios; and predicting test difficulty; Generate candidate detection solutions based on the body data, the analysis results of the detection target and the detection adaptability evaluation model, including: listing optional image detection methods; configuring detection parameter suggestions; designing detection timing arrangements; estimating detection costs; Optimize the selection of execution plans, including: evaluate the feasibility of each plan; compare the expected test results; weigh the cost-effectiveness of the test; and evaluate the acceptance of the subjects; Output the final detection plan as an auxiliary image detection plan, including: determining the main detection method; formulating alternative detection plans; generating detection precautions; and compiling detection process instructions.

6. The medical data intelligent analysis method based on infrared images according to claim 5 is characterized in that: The step of analyzing the auxiliary image data using a preset auxiliary image analysis model to obtain an auxiliary image analysis result includes: Load the preset auxiliary image analysis model, including: image preprocessing sub-model; organ segmentation sub-model; abnormality recognition sub-model; abnormality analysis sub-model; Performing image preprocessing on the auxiliary image data using the image preprocessing submodel to obtain first auxiliary image data, including: removing image noise; enhancing image contrast; correcting image distortion; and unifying image specifications; Using the organ segmentation sub-model and the first auxiliary image data to perform organ and tissue segmentation to obtain second auxiliary image data, including: identifying anatomical structures; segmenting target areas; extracting boundary information; marking key parts; Performing abnormal feature analysis using the abnormality recognition sub-model and the second auxiliary image data to obtain third auxiliary image data, including: detecting abnormal areas; extracting lesion features; measuring lesion sizes; and evaluating lesion properties; Use the abnormal analysis sub-model and the third auxiliary imaging data to perform multi-dimensional analysis, including: analyzing lesion distribution; evaluating lesion progression; comparing historical data; generating analysis suggestions; Output auxiliary image analysis results, including: organizing quantitative indicators; generating image annotations; summarizing analysis conclusions; and providing reference suggestions.

7. The medical data intelligent analysis method based on infrared images according to claim 6 is characterized in that: The step of obtaining a comprehensive analysis result based on the infrared image analysis result and the auxiliary image analysis result comprises: Performing data alignment and fusion on the infrared image analysis result and the auxiliary image analysis result, including: unifying the data format and coordinate system; registering the infrared and auxiliary image data; establishing the corresponding relationship of feature points; generating a fused data set; Construct a multimodal analysis model, including: loading the feature correlation model; configuring the analysis rule base; initializing the decision support system; setting weight parameters; Performing feature correlation analysis on the fused data set, including: extracting complementary feature information; analyzing feature correlation; identifying feature conflict points; synthesizing key features; Performing a comprehensive anomaly assessment based on the multimodal analysis model and the key features, including: integrating multi-source anomaly information; assessing the degree of evidence support; analyzing the reliability of anomaly assessment; and generating a comprehensive assessment opinion; Combined with comprehensive assessment opinions, perform risk assessment and forecasting, including: analyzing abnormal development trends and formulating risk warnings; Generate health analysis reports, including: summarizing key abnormal findings; generating multimodal images; providing health analysis recommendations; and formulating follow-up plans.

8. The medical data intelligent analysis method based on infrared images according to claim 7 is characterized in that: The method for constructing the infrared detection model includes: Build a quantum deep learning framework, including: designing the quantum convolution layer structure; building quantum pooling operations; defining quantum activation functions; configuring quantum fully connected layers; Designing a hybrid quantum-classical architecture, including: building a quantum feature extraction module; designing a classical neural network module; defining an inter-module interface protocol; and implementing a data conversion mechanism; Prepare training data sets, including: collecting infrared detection operation data, infrared detection equipment working data, infrared detection object data, and infrared detection working environment data; marking abnormal feature information; dividing data subsets; generating quantum state representations; Execute the model training process, including: initializing quantum parameters; performing quantum state evolution; calculating the loss function; updating model parameters; Implement model optimization to obtain the final infrared detection model, including: analyzing the impact of quantum noise; optimizing the depth of quantum circuits; adjusting hyperparameter configurations; and performing model pruning; Evaluate model performance, including: testing detection accuracy; evaluating computing resource consumption; analyzing model stability; and verifying generalization capabilities.

9. The medical data intelligent analysis method based on infrared images according to claim 8 is characterized in that: The method for constructing the infrared image analysis model includes: Build a federated learning network architecture, including: deploying a central coordination server; configuring local training nodes; establishing a secure communication channel; designing a data synchronization mechanism; Initialize the local analysis model, including: define the model structure; configure privacy protection parameters; set training hyperparameters; initialize the local data set; Perform federated model training, including: using collected infrared image sample data for local model training; encrypting model parameters; uploading parameter updates; aggregating global models; Implement differential privacy protection, including: calculating the privacy budget; adding noise perturbations; verifying the degree of privacy protection; adjusting protection parameters; Optimize model performance and obtain infrared image analysis model, including: evaluate model convergence; analyze parameter distribution; adjust aggregation strategy; optimize communication efficiency; Deploy and apply models, including: verifying model effects; deploying local services; monitoring operating status; and updating model versions.

10. A medical data intelligent analysis system based on infrared images, used to execute the medical data intelligent analysis method based on infrared images as claimed in any one of claims 1 to 9, characterized in that: include: A server, a first infrared detection terminal and a second infrared detection terminal; The server is configured to: Acquire first infrared image data collected by a first infrared detection terminal; Determining a detection target and a detection scheme of a second infrared detection terminal according to the first infrared image data and a preset infrared detection model; Controlling the second infrared detection terminal to perform detection according to the detection target and the detection scheme to obtain second infrared image data; Analyzing the second infrared image using a preset infrared image analysis model to obtain an infrared image analysis result; Determining an auxiliary imaging detection scheme according to the detection target and the physical data of the object being detected; Detecting the object to be detected according to the auxiliary image detection scheme to obtain auxiliary image data; Analyzing the auxiliary image data using a preset auxiliary image analysis model to obtain an auxiliary image analysis result; A comprehensive analysis result is obtained based on the infrared image analysis result and the auxiliary image analysis result.

Images