Chloasma diagnosis and treatment effect evaluation system based on multi-dimensional analysis

Abstract

The invention relates to the technical field of chloasma diagnosis and treatment effect evaluation, in particular to a chloasma diagnosis and treatment effect evaluation system based on multi-dimensional analysis. The system comprises an image segmentation processing module and a sample subset feature image selection module. According to the invention, the image segmentation processing module adopts a region-level image segmentation mode to carry out image segmentation processing on a collected lesion image, independent feature segmentation is carried out in combination with chloasma pathological features to obtain single feature images under different features, and meanwhile, sample subset feature images are marked through the sample subset feature image selection module, so that the sample subset feature images are obtained. And obtaining feature image screening items of each sample subset, updating features in a chloasma recovery stage in real time, and obtaining differences of the two features in combination with an initial lesion image to serve as a reference basis for evaluating the chloasma diagnosis and treatment effect, so that the chloasma diagnosis and treatment effect is evaluated in a multi-dimensional manner, the subjectivity of manual evaluation is reduced, and the evaluation efficiency is improved. Effects of different types of chloasma treatment schemes are objectively evaluated, and the evaluation accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of chloasma diagnosis and treatment effect evaluation, and in particular to a chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis. Background Art

[0002] Melasma is a common chronic facial hyperpigmentation skin disease characterized by symmetrically distributed light brown or dark brown patches on the face. The disease has an incidence rate of up to 30% among Asian women of childbearing age, and is prone to recurrence and difficult to cure, seriously affecting the patient's physical and mental health.

[0003] The formation of melasma is related to many factors, mainly including the following aspects: ① Endocrine changes: During pregnancy, the levels of estrogen and progesterone in the body increase, stimulating the proliferation of melanocytes and the synthesis of melanin, leading to the appearance of chloasma. After delivery, hormone levels return to normal and chloasma may be alleviated or disappear. Long-term oral contraceptives may also cause chloasma because the hormone components in contraceptives affect the activity of melanocytes.

[0004] ②Ultraviolet radiation: Long-term exposure to ultraviolet rays can stimulate the activity of melanocytes, leading to an increase in melanin and the formation of pigmentation. Chloasma is more common in exposed areas, with more severe skin lesions in summer and less severe in winter.

[0005] ③. Genetic factors: Chloasma may have a genetic tendency, and individuals with a family history of chloasma are more likely to suffer from the disease.

[0006] ④. Side effects of drugs: Certain drugs, such as anti-epileptic drugs and certain antipsychotic drugs, may cause the occurrence of chloasma by affecting the endocrine system or directly acting on melanocytes.

[0007] ⑤. Chronic diseases: Chronic diseases such as chronic hepatitis and tuberculosis will lead to increased tyrosinase activity and increased melanin production, while the body's own ability to eliminate melanin will be weakened. Over time, chloasma will occur on the face.

[0008] ⑥. Damage to the skin barrier: The skin barrier function declines, which in turn weakens the pigment barrier function, disrupts melanin metabolism, and causes melanin particles to deposit in the epidermis. For example, excessive cleaning or the use of highly irritating skin care products may damage the skin barrier, making the skin more vulnerable to ultraviolet damage and promoting the formation of chloasma.

[0009] Traditional treatment methods are mainly divided into two types: drug therapy and laser therapy. During the treatment process, images of patient lesions are collected at different stages, and image feature analysis is performed manually. The treatment effect of chloasma is evaluated based on the features. However, due to the different manifestation characteristics of chloasma in different types of patients and the strong subjectivity of manual evaluation, the evaluation effect is too single, and the feedback of treatment effects cannot be adapted to the diagnosis and treatment evaluation of different types of patients.

[0010] In order to address the above problems, a chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis is urgently needed. Summary of the invention

[0011] The purpose of the present invention is to provide a chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis to solve the problems raised in the above-mentioned background technology.

[0012] To achieve the above-mentioned purpose, a chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis is provided, which includes a chloasma patient lesion image acquisition module, an image segmentation processing module, a sample subset feature image selection module, and a recovery stage feature image update module; Wherein, the chloasma patient lesion image acquisition module is used to acquire lesion images of chloasma patients at different stages of recovery period; The image segmentation processing module uses a regional image segmentation method to perform image segmentation processing on the collected lesion image, and combines the pathological characteristics of chloasma to perform separate feature segmentation to obtain single feature images under different features; The sample subset feature image selection module obtains the minimized cut image in the process of regional level image segmentation, selects the minimized cut image with the highest coverage feature points, marks the sample subset feature image, and obtains the screening items of each sample subset feature image; The recovery stage feature image update module combines the sample subset feature images obtained at different stages and the corresponding screening items to update the features of the chloasma recovery stage in real time, and evaluates the chloasma treatment effect for non-disease diagnosis purposes in combination with the initial lesion image.

[0013] As a further improvement of the technical solution, the method for collecting lesion images of chloasma patients at different stages of the recovery period in the chloasma patient lesion image collection module comprises the following steps: S101, obtaining the pre-treatment lesion image and lesion appearance characteristics of the patient before treatment; S102, combining the appearance characteristics of the lesion image before treatment, obtaining the corresponding appearance characteristic state; S103, remove irrelevant features of the patient's lesion area; S104, periodically acquiring images of the patient's lesions using the same acquisition method.

[0014] As a further improvement of the technical solution, the image segmentation processing module includes an image feature point recognition unit, a feature value adjustment unit and a unit feature sub-image extraction unit; The image feature point recognition unit combines the collected lesion images at the treatment stage and extracts the feature points shown therein according to the pathological analysis; The feature value adjustment unit performs feature adjustment for region-level image segmentation according to the revealed feature points, obtains region-level images corresponding to different revealed feature points, and performs image segmentation processing to obtain single feature images under different features; The unit feature sub-image extraction unit performs sub-image feature value analysis on the segmented single feature images to obtain feature values ​​corresponding to each single feature image.

[0015] As a further improvement of the technical solution, the method for obtaining the feature values ​​corresponding to each single feature image in the unit feature sub-image extraction unit includes the following steps: S2301, collecting chloasma pathological data information, obtaining the color changes of chloasma, and formulating a color matching strategy; S2302, selecting a characteristic color range according to a color matching strategy as a characteristic point to be displayed; S2303: Obtain the color variation range of each sub-image in the single feature image, and use the color variation range with the highest color variation range ratio as the feature value corresponding to the current single feature image.

[0016] As a further improvement of the technical solution, the sample subset feature image screening items in the sample subset feature image selection module include plaque area, plaque shape, plaque arrangement and adjacent plaque interval; The patch area is the area formed by minimizing a single patch selected in the cut image; The patch shape is the shape corresponding to each patch selected in the minimized cut image; The patch arrangement method is to minimize the arrangement order of each patch selected in the cut image; The adjacent patch interval is to minimize the interval between adjacent patches selected in the cut image.

[0017] As a further improvement of the present technical solution, the sample subset feature image selection module includes a unit feature probability calculation unit and a calculation result screening unit. The unit feature probability calculation unit calculates the occurrence rate of the screening items in each minimized cutting image under the feature state to obtain the highest occurrence rate of each screening item. The calculation result screening unit marks the position of the corresponding minimized cutting image according to the highest occurrence rate of each screening item.

[0018] As a further improvement of the technical solution, the method for marking the position of the corresponding minimized cut image in the calculation result screening unit includes the following steps: S3201, obtaining the occurrence rate results of each screening item calculated in the unit feature probability calculation unit, and selecting the average value of each screening item corresponding to each minimized cut image; S3202, summing up the average values ​​of each screening item in each minimized cut image, performing secondary averaging processing, and obtaining the selected value of each screening item under the current feature; S3203, secondary matching of each minimized cut image, and taking the minimized cut image with the most selection values ​​that meet the screening items as the sample subset feature image under the current feature.

[0019] Compared with the prior art, the present invention has the following beneficial effects: In the chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis, the image segmentation processing module adopts the regional image segmentation method to perform image segmentation processing on the collected lesion images, and combines the pathological characteristics of chloasma to perform separate feature segmentation to obtain single feature images under different features. At the same time, the sample subset feature image selection module marks the sample subset feature image, and obtains the screening items of each sample subset feature image, updates the characteristics of the chloasma recovery stage in real time, and combines the initial lesion image to obtain the difference between the two as a reference for evaluating the diagnosis and treatment effect of chloasma, evaluates the diagnosis and treatment effect of chloasma in a multidimensional way, reduces the subjectivity of manual evaluation, objectively evaluates the effects of different types of chloasma treatment plans, and improves the evaluation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is the overall structural diagram of the present invention; Figure 2 A diagram showing the steps of a method for collecting images of lesions at different stages of the recovery period of chloasma patients according to the present invention; Figure 3 A diagram showing the steps of a method for obtaining feature values ​​corresponding to each single feature image according to the present invention; Figure 4 A method step diagram for marking the location of the corresponding minimized cut image of the present invention.

[0021] The meaning of each number in the figure is: 10. Chloasma patient lesion image acquisition module; 20. Image segmentation processing module; 210. Image feature point recognition unit; 220. Feature value adjustment unit; 230. Unit feature sub-image extraction unit; 30. Sample subset feature image selection module; 310. Unit feature probability calculation unit; 320. Calculation result screening unit; 40. Feature image update module in the restoration stage. DETAILED DESCRIPTION

[0022] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1 As shown, a chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis is provided, including a chloasma patient lesion image acquisition module 10, an image segmentation processing module 20, a sample subset feature image selection module 30 and a recovery stage feature image update module 40; Among them, the chloasma patient lesion image acquisition module 10 is used to acquire lesion images of chloasma patients at different stages of recovery period; The image segmentation processing module 20 uses a regional image segmentation method to perform image segmentation processing on the collected lesion image, and combines the pathological characteristics of chloasma to perform separate feature segmentation to obtain single feature images under different features; The sample subset feature image selection module 30 obtains the minimized cut image in the process of regional level image segmentation, selects the minimized cut image with the highest coverage of feature points, marks the sample subset feature image, and obtains the screening items of each sample subset feature image; The recovery stage characteristic image update module 40 combines the sample subset characteristic images obtained at different stages and the corresponding screening items to update the characteristics of the chloasma recovery stage in real time, and evaluates the chloasma treatment effect for non-disease diagnosis purposes in combination with the initial lesion image.

[0024] When used specifically, in the process of evaluating the diagnosis and treatment effect of chloasma, since the overall evaluation needs to be analyzed in combination with the patient's lesion images, it is first necessary to collect lesion images of chloasma patients at different stages of the recovery period through the chloasma patient lesion image acquisition module 10. The different stages referred to here are the treatment course stages during the treatment of the patient. For example, through laser treatment, one cycle of treatment is the initial stage of treatment. After completing one cycle of treatment, it is necessary to collect the patient's lesion image once. It is worth noting that the diagnosis and treatment data collected here are the diagnosis and treatment data of patients in the same treatment batch; After completing the lesion image acquisition, in order to further improve feature extraction, it is necessary to use the image segmentation processing module 20 to adopt the regional image segmentation method to perform image segmentation processing on the acquired lesion image, and combine the pathological characteristics of chloasma to perform separate feature segmentation to obtain single feature images under different features. This is because there are many characteristics that constitute chloasma. For example, in the active stage of chloasma, its characteristics are: the area of ​​lesions increases, the color deepens, the skin becomes red, the lesions become red after scratching, and most of the fragments fade when pressure is diagnosed; under RCM, there are more high-refractive, dendritic and stellate melanocytes in the basal layer of the epidermis with many and long dendrites, and the superficial dermis can be seen with varying numbers of medium-refractive mononuclear cell infiltration, and some high-refractive pigment-loving cells can be seen. Under the dermatoscope, the increase and expansion of capillaries can be seen.

[0025] During the stable stage of melasma, its characteristics are: the lesion area does not expand, the color does not deepen, the lesions do not turn red, the lesions do not turn red after scratching, and most of them do not fade when pressed on the slide; under RCM, there are fewer tree-shaped melanocytes in the basal layer of the epidermis, starburst-shaped melanocytes are relatively rare, and the number of mononuclear cells diffused in the superficial dermis is reduced.

[0026] Therefore, in the process of image segmentation, it is necessary to extract the pathological features of chloasma separately, and extract the single feature images under different features through regional level image segmentation. The specific steps are as follows: Step 1: Region growing: Select a seed point in the lesion image, and starting from the seed point, gradually add adjacent pixels into the region based on similarity; Step 2: Graph cutting: The lesion image is regarded as a whole image and the image is divided into different regions by minimizing the cutting cost. Step 3: Watershed algorithm: The image is regarded as terrain, and water is simulated to flow from high to low, so as to segment different areas. That is, based on the similar or identical factors of the regions, these factors are obtained manually and set as selection factors. These factors are the pathological characteristics of melasma. Each pathological characteristic of melasma corresponds to different factors. The model will perform regional selection according to the selection factors. The unselected part is the shadow part, and a single feature image corresponding to different pathological characteristics of melasma is generated. During the regional image segmentation process, multiple sub-images will be formed after the segmentation process, that is, in the above step 2, the image is divided into different regions by minimizing the cutting cost. At this time, the chloasma pathological features in each sub-image will be different, such as distribution rules, coverage area, and feature quantity and size. The sample subset feature image selection module 30 obtains the minimized cutting image in the regional image segmentation process, and selects the minimized cutting image with the highest coverage feature point, marks the sample subset feature image, and obtains the screening items of each sample subset feature image. Finally, the recovery stage feature image update module 40 combines the sample subset feature images obtained in different stages and the corresponding screening items to update the features of the chloasma recovery stage in real time, and combines the initial lesion image to obtain the difference between the two as a reference for evaluating the diagnosis and treatment effect of chloasma, that is, by analyzing the changes in the pathological features of each chloasma, the diagnosis and treatment effect of chloasma is evaluated in a multi-dimensional manner, the subjectivity of manual evaluation is reduced, and the effects of different types of chloasma treatment plans are objectively evaluated to improve the evaluation accuracy.

[0027] It is worth noting that the initial lesion image here is a characteristic image of a sample subset of the previous treatment stage, which further improves the comparison effect of the adjacent treatment stages and more intuitively evaluates the treatment effect.

[0028] In addition, if Figure 2 As shown, the method for collecting lesion images of chloasma patients at different stages of the recovery period in the chloasma patient lesion image collection module 10 includes the following steps: S101, obtaining the pre-treatment lesion image and lesion appearance characteristics of the patient before treatment; S102, combining the appearance characteristics of the lesion image before treatment, obtaining the corresponding appearance characteristic state; S103, remove irrelevant features of the patient's lesion area; S104, periodically acquiring images of the patient's lesions using the same acquisition method.

[0029] During specific use, during the process of collecting patient lesion images, in order to reduce the influence of irrelevant features, it is necessary to obtain the patient's pre-treatment lesion images and lesion manifestation characteristics in advance, and obtain the corresponding manifestation feature states, such as the coverage area and shape of chloasma, etc., which are all corresponding feature states. At the same time, different patients may have other lesion characteristics or skin characteristics in the lesion area, that is, accompanying features that are not chloasma, such as congenital freckles and scars left by other diseases, which are all non-correlated features. Finally, for later comparison, it is necessary to periodically collect patient lesion images using the same collection method, such as the range covered by the collected images and the camera angle during collection.

[0030] Furthermore, the image segmentation processing module 20 includes an image feature point recognition unit 210, a feature value adjustment unit 220, and a unit feature sub-image extraction unit 230; The image feature point recognition unit 210 combines the collected lesion images at the treatment stage and extracts the feature points shown therein according to the pathological analysis; The feature value adjustment unit 220 performs feature adjustment for regional level image segmentation according to the revealed feature points, obtains regional level images corresponding to different revealed feature points, and performs image segmentation processing to obtain single feature images under different features; The unit feature sub-image extraction unit 230 performs sub-image feature value analysis on the segmented single feature images to obtain feature values ​​corresponding to each single feature image.

[0031] Further, if Figure 3 As shown, the method for obtaining the feature value corresponding to each single feature image in the unit feature sub-image extraction unit 230 includes the following steps: S2301, collecting chloasma pathological data information, obtaining the color changes of chloasma, and formulating a color matching strategy; S2302, selecting a characteristic color range according to a color matching strategy as a characteristic point to be displayed; S2303: Obtain the color variation range of each sub-image in the single feature image, and use the color variation range with the highest color variation range ratio as the feature value corresponding to the current single feature image.

[0032] In specific use, since the symptoms of chloasma include not only changes in plaque shape and size, but also changes in color, according to clinical treatment, the color change range of chloasma can range from light yellow to dark brown. The darker the color, the more serious the condition and the higher the corresponding manifestation. Therefore, in the process of regional image segmentation, the feature division used is mainly distinguished by color change. First, the image feature point recognition unit 210 combines the collected lesion image at the treatment stage and extracts the feature points that appear therein according to pathological analysis, that is, the color change range of the plaque in the lesion area of ​​patients using the same treatment plan; Subsequently, the feature value adjustment unit 220 performs feature adjustment for regional-level image segmentation according to the displayed feature points, obtains regional-level images corresponding to different displayed feature points, and performs image segmentation processing to obtain single feature images under different features, and performs staged division according to the displayed color change range. Finally, the unit feature sub-image extraction unit 230 performs sub-image feature value analysis on the segmented single feature image to obtain the feature values ​​corresponding to each single feature image, that is, collects chloasma pathological data information, obtains the color change of chloasma, and formulates a color matching strategy. For example, in the lesion image of a patient, the color span of the plaque in the lesion area is dark yellow-dark brown. This is because the edge area of ​​the plaque is generally lighter in color than the central area, so there will be a certain color span, and this color span is the color span for later determination of chloasma. The main reference basis for the diagnosis and treatment effect of melasma. In order to further improve the evaluation accuracy, it is necessary to divide the color span of the plaques in one step and adjust them as features. For example, the color span of the above-mentioned plaques is dark yellow-dark brown, and the further divided features are dark yellow-light brown, light brown-middle brown and middle brown-dark brown. Each divided color range corresponds to a feature. The scanning reflectance spectrophotometer detection technology is used in the evaluation process: the Lab value of the plaque can be measured, where L represents the black and white brightness of the skin; a represents the red and green balance of the skin; b represents the yellow and blue balance of the skin, so as to evaluate the color of the chloasma plaque. The above-mentioned features finally obtain the color range, and this color range is used as the selection factor for regional selection. The unselected part is the shadow part, that is, the plaques in other color ranges are the shadow part; After completing the regional-level image segmentation of the single feature image, due to the different displayed colors of different sub-images in the spot color range, for example, in the regional-level image with the dark yellow-light brown color range as the feature, most areas in the selected sub-image appear light brown, and only a small amount of area appears dark yellow. At this time, the specific color light brown is used as the feature value corresponding to the current single feature image. Of course, there is still a color variation range in dark yellow-light brown, so the final selected feature is also a smaller color range, represented by the corresponding Lab value range.

[0033] Specifically, the sample subset feature image screening items in the sample subset feature image selection module 30 include plaque area, plaque shape, plaque arrangement and adjacent plaque interval; Among them, the patch area is the area formed by minimizing the single patch selected in the cut image; The patch shape is the shape corresponding to each patch selected in the minimized cut image; The patch arrangement method is to minimize the arrangement order of each patch selected in the cut image; The interval between adjacent patches is to minimize the interval between adjacent patches selected in the cut image.

[0034] In addition, the sample subset feature image selection module 30 includes a unit feature probability calculation unit 310 and a calculation result screening unit 320. The unit feature probability calculation unit 310 calculates the occurrence rate of the screening items in each minimized cutting image under the feature state to obtain the highest occurrence rate of each screening item. The calculation result screening unit 320 marks the position of the corresponding minimized cutting image according to the highest occurrence rate of each screening item.

[0035] Further, such as Figure 4 As shown, the method for marking the position of the corresponding minimized cut image in the calculation result screening unit 320 includes the following steps: S3201, obtaining the occurrence rate results of each screening item calculated by the unit feature probability calculation unit 310, and selecting the average value of each screening item corresponding to each minimized cut image; S3202, summing up the average values ​​of each screening item in each minimized cut image, performing secondary averaging processing, and obtaining the selected value of each screening item under the current feature; S3203, secondary matching of each minimized cut image, and taking the minimized cut image with the most selection values ​​that meet the screening items as the sample subset feature image under the current feature.

[0036] In the process of marking the sample subset feature image, since different types of chloasma involve different screening items, it is necessary to specify the screening items in advance. In the present invention, the screening items involved include plaque area, plaque shape, plaque arrangement and adjacent plaque interval. In order to improve the subsequent comparison effect, the unit feature probability calculation unit 310 is first required to calculate the occurrence rate of the screening items in each minimized cutting image under the feature state, and obtain the highest occurrence rate of each screening item. For example, a regional-level image with the feature of dark yellow-light brown forms n minimized cutting images. When calculating the screening item as plaque area, it is necessary to obtain the plaque area of ​​each plaque in a single minimized cutting image, and compare the plaque area therein. The patch area with the most identical data is marked as the selected area of ​​the current single minimized cutting image, that is, the patch area has the highest appearance rate. The patch shapes need to be compared by similarity, and the patch shape with the highest overlapping shape is used as the selected shape of the current single minimized cutting image, that is, the selected shape has the highest appearance rate. For the arrangement of patches, the arrangement of different minimized cutting images will be different, so each minimized cutting image has a specific patch arrangement. Finally, for the interval between adjacent patches, the length of each adjacent patch interval is obtained, and the final average value is calculated as the selected adjacent patch interval length of the current single minimized cutting image, that is, the adjacent patch interval length has the highest appearance rate. After completing the calculation of the occurrence rate of the screening items of a single minimized cutting image, the calculation result screening unit 320 marks the location of the corresponding minimized cutting image according to the highest occurrence rate of each screening item. Since there are a large number of minimized cutting images, it is necessary to first obtain the occurrence rate results of each screening item calculated in the unit feature probability calculation unit 310 during the specific marking, and select the average value of each screening item corresponding to each minimized cutting image, that is, taking the minimized cutting image as a whole, select the corresponding selection area, selection shape and selection of the interval length of adjacent patches. Since the feedback of the patch arrangement is wide and cannot be unified, this screening item The project is only used as a comparison project for the later sample subset feature image. Then the average values ​​of each screening item in each minimized cutting image are summarized, and secondary averaging is performed to obtain the selection values ​​of each screening item under the current feature. Finally, each minimized cutting image is matched twice, and the minimized cutting image that meets the most selection values ​​of the screening items is used as the sample subset feature image under the current feature. It is worth noting that in the specific matching process, a minimized cutting image may have multiple selection values ​​of screening items at the same time, or a selection value of a screening item may correspond to a minimized cutting image. Therefore, the number of sample subset feature images finally obtained will be different.

[0037] In the specific use process of the present invention, firstly, multiple image segmentation processes are performed using color range as a feature to obtain multiple groups of regional-level images under different features, and each group of regional-level images will obtain the color change range of each sub-image in the single feature image after segmentation processing, and use the color change range with the highest color change range ratio as the feature value corresponding to the current single feature image, as the comparison value under different features in the current stage, obtain the comparison value under different features in the next treatment cycle, perform feature value comparison of the same feature, obtain comparison results of different feature levels, and obtain differences in a more detailed manner. At the same time, the sample subset feature image selection module 30 marks the sample subset feature image during the image segmentation process, and obtains the screening items of each sample subset feature image. The sample subset feature images of different stages and the corresponding screening items are used to feedback the chloasma diagnosis and treatment status of the corresponding stage, and a secondary comparison is performed with the next treatment cycle to open up the direction of evaluating the diagnosis and treatment effect of chloasma in the later stage, and form a multidimensional numerical comparison result.

[0038] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis, characterized by: It comprises a chloasma patient lesion image acquisition module (10), an image segmentation processing module (20), a sample subset feature image selection module (30) and a recovery phase feature image update module (40); The chloasma patient lesion image acquisition module (10) is used to acquire lesion images of chloasma patients at different stages of the recovery period; The image segmentation processing module (20) uses a regional image segmentation method to perform image segmentation processing on the collected lesion image, and combines the pathological characteristics of chloasma to perform individual feature segmentation to obtain single feature images under different features; The sample subset feature image selection module (30) obtains the minimized cut image in the process of regional level image segmentation, selects the minimized cut image with the highest coverage of feature points, marks the sample subset feature image, and obtains the screening items of each sample subset feature image; The recovery stage characteristic image updating module (40) combines the sample subset characteristic images obtained at different stages and the corresponding screening items to update the characteristics of the chloasma recovery stage in real time, and evaluates the chloasma treatment effect for non-disease diagnosis purposes in combination with the initial lesion image.

2. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 1, characterized in that: The method for collecting images of lesions of chloasma patients at different stages of the recovery period in the chloasma patient lesion image collection module (10) comprises the following steps: S101, obtaining the pre-treatment lesion image and lesion appearance characteristics of the patient before treatment; S102, combining the appearance characteristics of the lesion image before treatment, obtaining the corresponding appearance characteristic state; S103, remove irrelevant features of the patient's lesion area; S104, periodically acquiring images of the patient's lesions using the same acquisition method.

3. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 1, characterized in that: The image segmentation processing module (20) comprises an image feature point recognition unit (210), a feature value adjustment unit (220) and a unit feature sub-image extraction unit (230); The image feature point recognition unit (210) combines the collected lesion image at the treatment stage and extracts the feature points shown therein according to the pathological analysis; The feature value adjustment unit (220) performs feature adjustment for region-level image segmentation according to the displayed feature points, obtains region-level images corresponding to different displayed feature points, and performs image segmentation processing to obtain single feature images under different features; The unit feature sub-image extraction unit (230) performs sub-image feature value analysis on the segmented single feature image to obtain feature values ​​corresponding to each single feature image.

4. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 3, characterized in that: The method for obtaining the characteristic values ​​corresponding to each single characteristic image in the unit characteristic sub-image extraction unit (230) comprises the following steps: S2301, collecting chloasma pathological data information, obtaining the color changes of chloasma, and formulating a color matching strategy; S2302, selecting a characteristic color range according to a color matching strategy as a characteristic point to be displayed; S2303: Obtain the color variation range of each sub-image in the single feature image, and use the color variation range with the highest color variation range ratio as the feature value corresponding to the current single feature image.

5. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 1, characterized in that: The sample subset characteristic image screening items in the sample subset characteristic image selection module (30) include plaque area, plaque shape, plaque arrangement and adjacent plaque interval; The patch area is the area formed by minimizing a single patch selected in the cut image; The patch shape is the shape corresponding to each patch selected in the minimized cut image; The patch arrangement method is to minimize the arrangement order of each patch selected in the cut image; The adjacent patch interval is to minimize the interval between adjacent patches selected in the cut image.

6. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 5, characterized in that: The sample subset feature image selection module (30) includes a unit feature probability calculation unit (310) and a calculation result screening unit (320). The unit feature probability calculation unit (310) calculates the occurrence rate of the screening items in each minimized cutting image under the feature state to obtain the highest occurrence rate of each screening item. The calculation result screening unit (320) marks the location of the corresponding minimized cutting image according to the highest occurrence rate of each screening item.

7. The chloasma diagnosis and treatment effect evaluation system based on multidimensional analysis according to claim 6, characterized in that: The method for marking the position of the corresponding minimized cut image in the calculation result screening unit (320) comprises the following steps: S3201, obtaining the occurrence rate results of each screening item calculated in the unit feature probability calculation unit (310), and selecting the average value of each screening item corresponding to each minimized cut image; S3202, summing up the average values ​​of each screening item in each minimized cut image, performing secondary averaging processing, and obtaining the selected value of each screening item under the current feature; S3203, secondary matching of each minimized cut image, and taking the minimized cut image with the most selection values ​​that meet the screening items as the sample subset feature image under the current feature.

Images