Radioactive skin injury monitoring and early warning analysis method and device based on multi-dimensional data

Abstract

The invention discloses a radioactive skin injury monitoring and early warning analysis method and device based on multi-dimensional data. The method comprises the steps that radioactive skin injury target block image data and corresponding natural data of a patient are collected in real time; a patient skin evaluation function is constructed according to the collected data, and the comprehensive unhealthy degree of the patient skin state is calculated; constructing a patient radioactive skin injury target block early warning function, and calculating a patient skin risk coefficient; on the basis of the radioactive skin injury stability performance description, according to patient target block data and comprehensive unhealth degree which are collected in real time, acquiring patient historical storage data, calculating a patient skin early warning factor of the comprehensive patient historical storage data, and generating a current state description of an injury target block; feedback description of the patient is obtained, and current state description of the damaged target block is updated. According to the method, positive feedback information beneficial to current judgment and nursing is obtained through recorded state recovery or deterioration conditions of similar conditions of the patient, and accurate intervention on the patient is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of early warning management, and in particular to a method and device for monitoring and early warning analysis of radiation skin damage based on multidimensional data. Background Art

[0002] With the widespread use of radiotherapy in oncology clinical care, the incidence of radiation-induced skin injury (RSI) in patients is as high as 87%-95%, exceeding 90% in patients with head and neck cancers, and can rapidly progress to severe lesions. RSI not only causes local symptoms such as pain, exudation, and secondary infection, but can also lead to ulceration and even necessitate treatment interruption in severe cases. It also significantly reduces patients' quality of life, increases psychological burden, prolongs hospital stays, and increases nursing workload and medical costs. Currently, clinical monitoring of RSI relies primarily on visual assessment by nursing staff and single biomarker testing: 1. Visual assessment is highly subjective, often overlooking subtle early lesions and making it difficult to guide personalized care in a timely manner; 2. Single biosensors focus on inflammatory factors or oxidative stress indicators, lacking the ability to comprehensively analyze multidimensional physiological and nursing indicators such as skin temperature, blood flow, and tissue elasticity; 3. Follow-up monitoring: Traditional follow-up intervals are long, and point-of-care evaluation and treatment plan adjustments lag, making it difficult to meet the rapidly changing needs of lesion intervention.

[0003] While artificial intelligence image analysis and big data technologies offer new insights into early microlesion identification, existing algorithms often rely on small sample sizes and fail to fully integrate real-time nursing assessment data and long-term follow-up records, making it difficult to achieve precise nursing early warning and dynamic intervention for radiation-induced skin injuries. Therefore, a multidimensional data-based radiation skin injury monitoring and early warning analysis method and device are urgently needed to address these issues. Summary of the Invention

[0004] In order to solve the above-mentioned deficiencies in the prior art, the present invention aims to overcome the existing deficiencies and provide a method for monitoring and early warning analysis of radiation skin damage based on multidimensional data, comprising the following: Real-time acquisition of target block image data of radiation-induced skin damage of the patient and natural data corresponding to the target block image data of radiation-induced skin damage of the patient; Based on the collected radiation skin damage target block image data and the natural data of the corresponding period, a patient skin evaluation function is constructed to calculate the comprehensive unhealthiness of the patient's skin condition; Based on the generated comprehensive unhealthiness of the patient's skin condition, a warning function for the patient's radiation skin damage target block is constructed, the patient's skin risk coefficient is calculated, and a description of the stability performance of radiation skin damage is generated; Based on the generated stable performance description of radiation skin damage, according to the real-time collection of the patient's radiation skin damage target block data and the comprehensive unhealthiness of the patient's skin condition, the patient's historical storage data is obtained, and the patient's skin warning factor based on the comprehensive patient's historical storage data is calculated to generate a current state description of the damage target block; Obtain real-time feedback description of the patient's radiation skin damage target block treatment, and update the current status description of the damage target block based on the patient's feedback description.

[0005] As a further optimization of the above solution, the real-time collected target block image data of the radiation-induced skin lesions of the patient is analyzed, including the following: Acquire image data of patients undergoing radiation skin injury treatment, and perform regional segmentation on the acquired image data into multiple sub-target blocks of equal area; According to the skin damage monitoring results of the target block data images of radioactive skin damage of historical patients, the multiple segmented sub-target blocks of equal area are re-labeled and represented as high-frequency skin damage intervals and low-frequency skin damage intervals respectively; Continue to acquire image data of the current patient undergoing radiation skin injury treatment, acquire images to be compared with the current patient, compare the real-time acquired image data with the images to be compared, and obtain an interval to be identified of the image data of the current patient undergoing radiation skin injury treatment; According to the historical image marking results of the high-frequency skin damage interval and the low-frequency skin damage interval, the interval to be identified of the current patient is marked as a high-frequency skin damage interval sub-target block or a low-frequency skin damage interval sub-target block according to the high-frequency skin damage interval and the low-frequency skin damage interval, and skin damage identification analysis is performed on the analyzed sub-target blocks; Based on the skin damage identification analysis of the recorded high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block, the confirmed damage image area of ​​the radioactive skin damage is obtained. According to the confirmed damage image area of ​​the radioactive skin damage, it is confirmed whether the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block is a real radioactive skin damage area, and the real area position and size are recorded.

[0006] As a further optimization of the above solution, the specific method of constructing the patient skin evaluation function and calculating the comprehensive unhealthiness of the patient's skin condition includes the following: Based on the collected radiation skin damage target block data and the natural data of the corresponding period, a patient skin evaluation function is constructed, including the following: In the above formula, For patients The overall unhealthiness of the skin condition under the collected data; Indicates the unhealthiness of the skin condition under a single collection of data; The number of times the target data for the patient is obtained; For the The target area data of radiation skin damage obtained by the first time; For the The natural data corresponding to the radiation skin damage target block data acquired; and are the radiation skin injury preparation parameters, As an influencing factor.

[0007] As a further optimization of the above scheme, the method for constructing a warning function for the target block of radiation-induced skin damage in patients includes the following: According to the constructed patient skin evaluation function, the patient skin risk coefficient is calculated: In the above formula, is the patient's skin risk factor, is the adjustment item for natural data, 、 and are the preset regression coefficients respectively.

[0008] As a further optimization of the above solution, the method further includes the following: According to the calculated patient skin warning probability, it is changed into a radiation skin damage stability performance description parameter and output accordingly, see the following formula: In the above formula, is the stability performance description parameter, is the threshold used to describe the stability performance, and They are Patient Skin Warning and Patient Skin Stable; According to the calculated stability performance description parameters, specific evaluation information of the radiation skin damage target block is sent accordingly.

[0009] As a further optimization of the above solution, the method for calculating the patient skin warning factor based on the patient's historical stored data to generate a description of the current state of the damaged target block includes the following: In the above formula, It is a patient skin warning factor based on the patient's historical storage data. Store data for patient history, , and are preset weighting parameters respectively; Preset the current patient's radiation skin damage target block status assessment threshold: if The value exceeds the maximum value of the current patient's radiation skin damage target block state assessment threshold, and the corresponding The parameters are , output normal level; if The value is lower than the minimum value of the target block state assessment threshold of the current patient's radiation skin damage, and the corresponding The parameters are , output high risk level; if If the value is between the target block status assessment thresholds for radiation skin damage of the current patient, a low risk level is output.

[0010] As a further optimization of the above solution, the method of obtaining a real-time feedback description of the radiation skin damage target block of the patient and updating the current state description of the damage target block based on the patient feedback description includes the following: Get the parameter information of the radiation skin damage target block of the current patient and calculate the updated radiation skin damage target block data fed back by the current patient: In the above formula, Updated radiation skin injury target block data for current patient feedback, Score the abnormal performance of target blocks of radiation-induced skin damage. To score the target mass sensation of radiation skin injury, and are the weighting coefficients of abnormal expression score and somatosensory score, respectively.

[0011] As a further optimization of the above solution, the method further includes the following: Based on the updated radiation skin damage target block data and the patient's historical storage data, the current status description of the damage target block is updated: in, The estimated difference between the current patient's radiation skin damage target block status and the patient's historical storage status; Preset the current patient's radiation skin damage target block state change threshold: If the calculated Exceeding the state change threshold of the radiation skin damage target block of the current patient indicates that the radiation skin damage target block of the current patient has changed significantly, and the current state description of the damage target block is updated.

[0012] The present invention discloses a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein when the computer program instructions are executed by the processor, the steps of the method for monitoring and early warning analysis of radioactive skin damage based on multidimensional data as described in the present invention are implemented.

[0013] The present invention discloses a computer-readable storage medium having a computer program stored thereon, wherein the computer program instructions, when executed by a processor, implement the steps of the radioactive skin damage monitoring and early warning analysis method based on multidimensional data of the present invention.

[0014] The present invention adopts the above technical solution. Compared with the existing technology, a method and device for monitoring and early warning analysis of radiation skin damage based on multidimensional data has the following technical effects: 1. In the present invention, by constructing a patient skin assessment function and combining the radiation skin damage target block data with the natural data corresponding to the patient's collected target block data, effective radiation skin damage target block record information of any state or time is obtained, which can be used to predict various subsequent skin recovery situations and provide clinical nurses with a scientific decision-making basis.

[0015] 2. The present invention calculates the result prompts of the radiation skin damage target block warning and transmits accurate radiation skin damage target block protection prompts, which can make patients know the progress of radiation skin damage recovery or the damage risk level in advance, and simultaneously push the grading results and personalized nursing plans to the nurse workstation, helping the nursing team to intervene in time before the symptoms worsen.

[0016] 3. The present invention uses the patient's historical storage data and the current patient's radiation skin damage target block data to continuously call the historical nursing records and follow-up results of similar cases, and incorporates the current patient's feedback data into the technical solution update to achieve dynamic correction of skin damage warnings and nursing plans, thereby improving nursing quality and patient compliance. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Other features, objects and advantages of the present application will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a schematic diagram of the process of the present invention; Figure 2 This is another schematic flow chart of the present invention. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] Example 1: like Figure 1As shown, the embodiment of the present invention discloses a method for monitoring and early warning analysis of radiation skin damage based on multidimensional data, including the following: Step 1: Real-time acquisition of target block image data of the patient's radioactive skin lesion and natural data corresponding to the target block image data of the patient's radioactive skin lesion. Specifically, the target block image data of the radioactive skin lesion of the present invention is the area of ​​the target block of the patient's radioactive skin lesion acquired by any electronic image acquisition device, and the corresponding natural data is the influencing parameters under the natural state corresponding to the acquisition of the target block area of ​​the patient's radioactive skin lesion, such as air temperature and humidity, to form a complete target block data set of the radioactive skin lesion. Specifically, the real-time collected target block image data of radiation-induced skin damage of patients is analyzed, including the following: Step 11: Acquire image data of the patient undergoing radiation-induced skin injury treatment, and perform regional segmentation on the acquired image data into a plurality of sub-target blocks of equal area. It should be noted that the image data for radiation-induced skin injury treatment in the present invention is acquired from the complete radioactively affected area of ​​the patient, and the regional segmentation is actually performed using a predetermined segmentation scheme or by medical professionals. Step 12: Based on the skin damage monitoring status of the radiation-induced skin damage target block image data of the historical patients, the multiple segmented sub-target blocks of equal area are re-labeled to represent high-frequency skin damage intervals and low-frequency skin damage intervals respectively; More specifically, to establish a scientific method for monitoring and early warning of radiation-induced skin damage, the present invention incorporates target block image data of radiation-induced skin damage from historical patients. This includes image data of radiation-induced skin damage acquired during any treatment process prior to the current treatment, whether for the same patient or for other patients. Of particular note, image feature extraction and matching algorithms are used to fit features of the skin lesion areas of different patients, allowing the acquired images to be treated as highly similar "equivalent samples," enhancing the generalization and prediction accuracy of the present invention, similar to image acquisition of the same human tissue site.

[0020] More specifically, the division method of high-frequency skin damage interval and low-frequency skin damage interval is as follows: For any patient to be treated, the sampling period is one day, and the location information of the radiation skin damage after a few days is recorded, which is marked as a sampling stage. ,The specific collection days are adjusted adaptively according to different treatment plans; The frequency of radiation-induced skin damage in the segmented sub-regions during a sampling period is recorded as , let the total number of segmented sub-regions be , then ,in, For any sub-region; calculate , for any sub-region frequency description function If the numerical value exceeds the preset description threshold, the corresponding sub-region is marked as a high-frequency skin damage interval, otherwise it is a low-frequency skin damage interval.

[0021] Step 13, continue to acquire image data of the current patient undergoing radiation skin injury treatment, collect the current patient's image to be compared, compare the real-time acquired image data with the image to be compared, and obtain the interval to be identified of the image data of the current patient undergoing radiation skin injury treatment; More specifically, the image recorded in the present invention can be compared with the image to be compared by performing a comparison through image feature point analysis or image pixel concentration, thereby quickly obtaining a preliminary judgment result of the image to be compared and confirming whether further analysis is required and the area to be analyzed. It should be particularly noted that the image to be compared in the present invention is specifically represented by an image of a person in a normal state without damage, and the location of the image is the same as the location where radiation skin damage treatment is performed. For example, an image of the inner arm between the palm and the first elbow joint of a patient in a normal state is the image to be compared, and an image taken after radiation skin damage occurs in this area is the image of the radiation skin damage treatment. Step 14: Based on the historical image labeling results of the high-frequency skin lesion interval and the low-frequency skin lesion interval, the interval to be identified of the current patient is labeled as a high-frequency skin lesion interval sub-target block or a low-frequency skin lesion interval sub-target block according to the high-frequency skin lesion interval and the low-frequency skin lesion interval, and skin lesion identification analysis is performed on the analyzed sub-target blocks; More specifically, the analysis of skin lesions includes the following: Select any high-frequency skin damage interval sub-target block or low-frequency skin damage interval sub-target block, use it as the description reference, and set the first description surface based on its plane position; based on the first description surface, move the height toward the outside of the skin. The second description plane is set based on the distance between the two locations. It should be noted that, although the local location of human tissue can be regarded as a plane, there are still tissue locations with large differences. Therefore, when selecting the description benchmark, it is necessary to consider the skin damage location as much as possible, in order to select the location with the largest approximate flat area. According to the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block marked above, distance anchor lines are drawn on the second description surface in the direction of the first description surface; specifically, the anchor line position is within the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block, and the number of distance anchor lines drawn in the corresponding sub-target block is adaptively adjusted according to the high frequency or low frequency of the sub-target block, specifically, the number of distance anchor lines corresponding to the high-frequency skin damage interval sub-target block range is higher than the number of distance anchor lines corresponding to the low-frequency skin damage interval sub-target block range, and the specific number is selected according to the actual condition of the disease; Count the lengths of all distance anchor lines , calculate all distance anchor line lengths and moving heights The absolute value of the difference is recorded as ; Indicates the accumulation of all One of the distance anchor lines, then It should be noted that the present invention It is mainly determined by the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block. D is represented by the distance anchor line drawn within any high-frequency skin damage interval sub-target block or low-frequency skin damage interval sub-target block. The value of D is determined by professionals. The larger the D is in the high-frequency skin damage interval sub-target block, the smaller the D is in the low-frequency skin damage interval sub-target block. The absolute value of the difference Describing Thresholds for Radiation-Injured Skin For comparison, if , indicating that the difference between the two endpoint planes is large, which is an obvious skin injury; otherwise, it means that the two endpoint planes are close, which is an inconspicuous skin injury; According to the statistical results of the absolute value of the difference and the description threshold of radiation skin damage, the number of marked obvious skin damage is obtained and recorded as ,calculate exist If the proportion value is greater than the identification threshold , then all intersections of the distance anchor lines marked as obvious skin damage and the high-frequency skin damage interval sub-target blocks or low-frequency skin damage interval sub-target blocks are recorded as skin damage points; The present invention performs skin lesion identification and analysis on sub-target blocks and adopts the distance relationship between the distance anchor line and the description reference plane to obtain the identification results of all radioactive skin lesions of the patient, thereby improving the accuracy of the final injury identification results. At the same time, a radioactive skin lesion description threshold is set to accurately circle the skin location of the skin lesion that actually needs treatment, remove interference items, improve identification efficiency, and reduce computational redundancy for subsequent treatment plans.

[0022] Step 15, based on the skin damage identification analysis of the recorded high-frequency skin damage interval sub-target block or low-frequency skin damage interval sub-target block, obtain the confirmed damage image area of ​​the radioactive skin damage, confirm whether the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block is a real radioactive skin damage area according to the confirmed damage image area of ​​the radioactive skin damage, and record the real area position and size; based on this technology, the radioactive skin damage target block image data can be accurately obtained, and the area formed by the skin damage points is the confirmed damage image area of ​​the radioactive skin damage.

[0023] Step 2: Based on the collected radiation skin damage target block image data and the natural data of the corresponding time period, a patient skin evaluation function is constructed to calculate the comprehensive unhealthiness of the patient's skin condition; Based on the collected radiation skin damage target block data and the natural data of the corresponding period, a patient skin evaluation function is constructed, including the following: In the above formula, For patients The overall unhealthiness of the skin condition under the collected data; Indicates the unhealthiness of the skin condition under a single collection of data; The number of times the target data for the patient is obtained; For the The target area data of radiation skin damage obtained by the first time; For the The natural data corresponding to the radiation skin damage target block data acquired; and are the radiation skin injury preparation parameters, As an influencing factor.

[0024] Function is constructed by The relationship between the target area of ​​radioactive skin damage and the corresponding natural environment data is combined to describe the current skin health status of the patient in digital form: when the target area of ​​radioactive skin damage is small, the unhealthy degree of the patient's skin is mainly determined by the collected environmental data. For example, taking temperature as an example, the higher the temperature, the better the skin health. The larger the value, the more unhealthy the patient's skin is; when the ambient temperature data approaches 0℃, The function becomes a linear function, and the area of ​​the target patch of radiation-induced skin damage directly determines the degree of unhealthy skin. When the temperature is high, the degree of unhealthy skin is determined by both the area of ​​the target patch of radiation-induced skin damage and the temperature of the acquisition environment. If the temperature is low, the influence of the acquisition environment temperature exceeds that of the area of ​​the target patch of radiation-induced skin damage. and Prepare parameters for radiation skin damage respectively, and adjust the weights of radiation skin damage target block data and natural data on the results respectively. is an influencing factor used to control the influence of the exponential decay term.

[0025] In the present invention, by constructing a patient skin evaluation function, based on the combined radiation skin damage target block data and the natural data corresponding to the patient's collected target block data, effective radiation skin damage target block record information of any state or time is obtained, which can be used for the subsequent prediction of various skin recovery situations.

[0026] Step 3: Based on the generated comprehensive unhealthiness of the patient's skin condition, a warning function for the patient's radiation skin damage target block is constructed, the patient's skin risk coefficient is calculated, and a description of the stability performance of the radiation skin damage is generated; Specifically, the method for constructing a warning function for a target block of radiation-induced skin damage in patients includes the following: According to the constructed patient skin evaluation function, the patient skin risk coefficient is calculated: In the above formula, is the patient's skin risk factor, This is an adjustment item for natural data, which indicates the degree of influence of the natural environment on the recovery or deterioration of the patient's skin. 、 and are the preset regression coefficients, where and Respectively represent the weights of the patient skin assessment description value and the adjustment items of natural data on the patient skin warning; It should be noted that, generally speaking, the present invention The range value is between 0 and 1. The closer it is to 1, the higher the risk of skin warning, that is, the condition may worsen. The present invention combines the comprehensive unhealthy degree of skin condition with the adjustment item of natural data to construct a linear relationship. and The value is a positive number, so the patient's skin risk coefficient is positively correlated with the comprehensive unhealthiness of the skin condition and the adjustment item of the natural data. That is, the higher the comprehensive unhealthiness of the skin condition and the more obvious the adjustment of the natural data, the higher the probability of skin risk warning for the patient.

[0027] More specifically, based on the calculated patient skin warning probability, it is changed into a radiation skin damage stability performance description parameter and output accordingly, as shown in the following formula: In the above formula, is the stability performance description parameter, is the threshold used to describe the stability performance, and They are respectively patient skin warning and patient skin stability; when special instructions are required, generally speaking, the present invention The value is 0.6, that is, if the calculated A value of 0.6 or less indicates low risk; According to the calculated stability performance description parameters, specific evaluation information of the radiation skin damage target block is sent accordingly.

[0028] The present invention calculates the result prompt of the radiation skin damage target block warning and transmits accurate radiation skin damage target block protection prompt, which can make patients know the progress of radiation skin damage recovery or deterioration in advance, so as to facilitate early response; Step 4, based on the generated stable performance description of the radiation skin damage, according to the real-time collection of the patient's radiation skin damage target block data and the comprehensive unhealthiness of the patient's skin status, the patient's historical storage data is obtained, and the patient's skin warning factor based on the comprehensive patient's historical storage data is calculated to generate a current status description of the damage target block; the present invention uses the patient's historical storage data and the radiation skin damage target block data of the current patient, and thus obtains positive feedback information that is beneficial to current judgment and care through the status recovery or deterioration of similar patients that have been recorded, which is conducive to the correctness of patient selection.

[0029] Specifically, the method for generating the current state description of the damaged target block includes the following: In the above formula, It is a patient skin warning factor based on the patient's historical storage data. Store data for patient history, , and are preset weighting parameters respectively; Set the current patient's radiation skin damage target block status assessment threshold in advance: if The value exceeds the maximum value of the current patient's radiation skin damage target block state assessment threshold, and the corresponding The parameters are , output normal level; if The value is lower than the minimum value of the target block state assessment threshold of the current patient's radiation skin damage, and the corresponding The parameters are , output high risk level; if The value is between the current patient radiation skin damage target block state assessment threshold, and the output is a low risk level; More specifically, in this technology, it is determined that the user's patient skin stable state is , which means that the patient is not at high risk at present, combined with the radiation skin damage target block warning parameters When it is higher than the maximum warning threshold, it indicates that the patient's current skin condition is stable and conventional care is sufficient. Otherwise, emergency care and other related measures are required. It is particularly important to note that It is used to feedback the warning level of the patient's radiation skin damage target block. The value is the most recently acquired target block image of the patient's radiation skin damage and the corresponding natural environment data; The patient history storage data is specifically represented as descriptive data generated based on the historical patient information reported by the same patient. The values ​​are selected based on a fixed period and determined by medical staff. Step 5: Obtain the real-time feedback description of the patient's radiation skin damage target block treatment, and update the current status description of the damage target block based on the patient's feedback description.

[0030] Specifically, a method for obtaining a real-time feedback description of the radiation skin damage target block of a patient and updating the current state description of the damage target block based on the patient feedback description includes the following: Get the parameter information of the radiation skin damage target block of the current patient and calculate the updated radiation skin damage target block data fed back by the current patient: In the above formula, is the comprehensive evaluation value after the current patient feedback, Score the abnormal performance of the target area of ​​radiation skin damage, such as the corresponding evaluation value of the itching, severe pain, cooling and other experiences that may occur during the recovery or treatment of radiation skin damage. To provide a somatosensory score for the target area of ​​radiation skin damage after treatment, such as feedback on the diagnosis, treatment and care of the recovery process of radiation skin damage, and are the weighted coefficients of the abnormal performance score and the somatosensory score, respectively, indicating the degree of influence of the abnormal performance score and the somatosensory score on the patient's global feedback; more specifically, Indicates the real-time feedback information of the patient's radiation-induced skin damage, which includes both objective information about the patient's damaged skin and a description of the state after recovery or treatment; More specifically, based on the calculated updated radiation skin damage target block data fed back by the current patient and the patient's historical stored data, the current state description of the damage target block is updated: in, The estimated difference between the current patient's radiation skin damage target block status and the patient's historical storage status; Preset the current patient's radiation skin damage target block state change threshold: If the calculated Exceeding the threshold for the current patient's radiation-induced skin injury target block state change indicates a significant change in the target block, and the current state description of the damage target block is updated. Specifically, based on the constructed comprehensive assessment value and the patient's historical stored data, the present invention can quickly and timely update the current state description of the damage target block based on the patient's current state and historical records. This enables dynamic tracking and monitoring of radiation-induced skin injury symptoms in modern medical care, and allows for more accurate recovery or treatment plans based on user status feedback.

[0031] The present invention realizes integrated modeling of multi-source data by jointly analyzing image data of radiation skin damage areas with individual patient influencing factors and natural environmental parameters, thereby enhancing the objectivity and accuracy of the analysis system and improving the reliability of subsequent damage risk prediction.

[0032] Example 2: The present invention provides another example to verify the correctness of the beneficial effects of the present invention. 50 patients with radiation-induced skin damage were selected and their changes over 15 days were recorded in detail: Early stage: The early stage of the present invention specifically refers to the stage of collecting the radiation skin damage target block data of the patient and the natural data corresponding to the collected target block data of the patient, and the recording is for 5 days; each person collects data once every hour from 8:00 to 22:00 every day, and finally records the average value as the collected data of any patient on that day; Mid-term: The mid-term of the present invention specifically refers to the stage of generating patient skin assessment prompts, which is recorded as 5 days; Late stage: The late stage of the present invention specifically indicates the stage of early warning of the target block of radiation skin damage of the current patient, which is recorded as 5 days; According to the above table, and The two parameters showed a negative correlation trend. The value is 77, but its corresponding Very low; Patient F The value is 62, but its corresponding Very high. Through this control group, it can be analyzed that when the natural data is poor, that is, the temperature is too high above the normal temperature, the recovery of the target block of radiation skin damage is poor; Analysis of patient F The value of 0.64 is analyzed, which is the highest value of all recorded patients. Marked as H, indicating that the patient needs intensive care for radiation skin injury.

[0033] The above description is merely a preferred embodiment of the present application and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the inventive concept. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for monitoring and early warning of radiation skin damage based on multidimensional data, characterized in that: The method comprises the following: Real-time acquisition of target block image data of radiation-induced skin damage of the patient and natural data corresponding to the target block image data of radiation-induced skin damage of the patient; Based on the collected radiation skin damage target block image data and the natural data of the corresponding period, a patient skin evaluation function is constructed to calculate the comprehensive unhealthiness of the patient's skin condition; Based on the generated comprehensive unhealthiness of the patient's skin condition, a warning function for the patient's radiation skin damage target block is constructed, the patient's skin risk coefficient is calculated, and a description of the stability performance of radiation skin damage is generated; Based on the generated stable performance description of radiation skin damage, according to the real-time collection of the patient's radiation skin damage target block data and the comprehensive unhealthiness of the patient's skin condition, the patient's historical storage data is obtained, and the patient's skin warning factor based on the comprehensive patient's historical storage data is calculated to generate a current state description of the damage target block; Obtain real-time feedback description of the patient's radiation skin damage target block treatment, and update the current status description of the damage target block based on the patient's feedback description.

2. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 1, characterized in that: Analyzing the real-time collected target block image data of the radiation-induced skin lesion of the patient includes the following steps: Acquire image data of patients undergoing radiation skin injury treatment, and perform regional segmentation on the acquired image data into multiple sub-target blocks of equal area; According to the skin damage monitoring results of the target block data images of radioactive skin damage of historical patients, the multiple segmented sub-target blocks of equal area are re-labeled and represented as high-frequency skin damage intervals and low-frequency skin damage intervals respectively; Continue to acquire image data of the current patient undergoing radiation skin injury treatment, acquire images to be compared with the current patient, compare the real-time acquired image data with the images to be compared, and obtain an interval to be identified of the image data of the current patient undergoing radiation skin injury treatment; According to the historical image marking results of the high-frequency skin damage interval and the low-frequency skin damage interval, the interval to be identified of the current patient is marked as a high-frequency skin damage interval sub-target block or a low-frequency skin damage interval sub-target block according to the high-frequency skin damage interval and the low-frequency skin damage interval, and skin damage identification analysis is performed on the analyzed sub-target blocks; Based on the skin damage identification analysis of the recorded high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block, the confirmed damage image area of ​​the radioactive skin damage is obtained. According to the confirmed damage image area of ​​the radioactive skin damage, it is confirmed whether the high-frequency skin damage interval sub-target block or the low-frequency skin damage interval sub-target block is a real radioactive skin damage area, and the real area position and size are recorded.

3. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 1 is characterized in that: The specific method of constructing the patient skin evaluation function and calculating the comprehensive unhealthiness of the patient's skin condition includes the following: Based on the collected radiation skin damage target block data and the natural data of the corresponding period, a patient skin evaluation function is constructed. In the above formula, For patients The overall unhealthiness of the skin condition under the collected data; Indicates the unhealthiness of the skin condition under a single collection of data; The number of times the target data for the patient is obtained; For the The target area data of radiation skin damage obtained by the first time; For the The natural data corresponding to the radiation skin damage target block data acquired; and are the radiation skin injury preparation parameters, As an influencing factor.

4. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 3 is characterized in that: The method for constructing the warning function of the target block of radiation skin damage of patients includes the following: According to the constructed patient skin evaluation function, the patient skin risk coefficient is calculated: In the above formula, is the patient's skin risk factor, is the adjustment item for natural data, 、 and are the preset regression coefficients respectively.

5. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 4 is characterized in that: The method further comprises: According to the calculated patient skin warning probability, it is changed into a radiation skin damage stability performance description parameter and output accordingly, see the following formula: In the above formula, is the stability performance description parameter, is the threshold used to describe the stability performance, and They are Patient Skin Warning and Patient Skin Stable; According to the calculated stability performance description parameters, specific evaluation information of the radiation skin damage target block is sent accordingly.

6. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 5, characterized in that: The method for calculating the patient skin warning factor based on the patient's historical stored data to generate a description of the current state of the damaged target block includes the following: In the above formula, It is a patient skin warning factor based on the patient's historical storage data. Store data for patient history, , and are preset weighting parameters respectively; Preset the current patient's radiation skin damage target block status assessment threshold: if The value exceeds the maximum value of the current patient's radiation skin damage target block state assessment threshold, and the corresponding The parameters are , output normal level; if The value is lower than the minimum value of the target block state assessment threshold of the current patient's radiation skin damage, and the corresponding The parameters are , output high risk level; if If the value is between the target block status assessment thresholds for radiation skin damage of the current patient, a low risk level is output.

7. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 6, characterized in that: The method of obtaining a real-time feedback description of the radiation skin damage target block of the patient and updating the current state description of the damage target block based on the patient feedback description includes the following: Get the parameter information of the radiation skin damage target block of the current patient and calculate the updated radiation skin damage target block data fed back by the current patient: In the above formula, Updated radiation skin injury target block data for current patient feedback, Score the abnormal performance of target blocks of radiation-induced skin damage. To score the target mass sensation of radiation skin injury, and are the weighting coefficients of abnormal expression score and somatosensory score, respectively.

8. The method for monitoring and early warning of radiation skin damage based on multidimensional data according to claim 7, characterized in that: The method further comprises: Based on the updated radiation skin damage target block data and the patient's historical storage data, the current status description of the damage target block is updated: in, The estimated difference between the current patient's radiation skin damage target block status and the patient's historical storage status; Preset the current patient's radiation skin damage target block state change threshold: If the calculated Exceeding the state change threshold of the radiation skin damage target block of the current patient indicates that the radiation skin damage target block of the current patient has changed significantly, and the current state description of the damage target block is updated.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the radiation skin damage monitoring and early warning analysis method based on multidimensional data are implemented as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for monitoring and early warning of radiation skin damage based on multidimensional data according to any one of claims 1 to 8 are implemented.

Images