CN110544534B - Method and system for automatically evaluating treatment effect of skin disease - Google Patents
Method and system for automatically evaluating treatment effect of skin disease Download PDFInfo
- Publication number
- CN110544534B CN110544534B CN201910813912.3A CN201910813912A CN110544534B CN 110544534 B CN110544534 B CN 110544534B CN 201910813912 A CN201910813912 A CN 201910813912A CN 110544534 B CN110544534 B CN 110544534B
- Authority
- CN
- China
- Prior art keywords
- focus
- treatment effect
- image
- skin disease
- focus image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/44—Detecting, measuring or recording for evaluating the integumentary system, e.g. skin, hair or nails
- A61B5/441—Skin evaluation, e.g. for skin disorder diagnosis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4848—Monitoring or testing the effects of treatment, e.g. of medication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30088—Skin; Dermal
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- Biophysics (AREA)
- Animal Behavior & Ethology (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Dermatology (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a method and a system for automatically evaluating the treatment effect of skin diseases. The automatic evaluation method for the dermatosis treatment effect can enable a patient to quickly obtain accurate treatment effect evaluation information according to the focus images which are regularly collected and uploaded, and the treatment effect evaluation information covers the area ratio of the dermatosis, the area ratio of focus grading (heavy, medium and light), a treatment effect timeline, medical advice and the like, so that the dermatosis patient can obtain accurate treatment information, the confidence of diagnosis and treatment schemes can be enhanced, the uncertainty is eliminated, and the relationship between doctors and patients is improved.
Description
Technical Field
The invention relates to a method and a system for automatically evaluating the treatment effect of skin diseases.
Background
Skin disorders (dermatasis) are a general term for disorders occurring in the skin and its appendages. The variety of skin disorders is wide, and common skin disorders include: viral, bacterial, fungal and allergic skin diseases. Identification of skin diseases and qualitative analysis of the processes is a huge effort. At present, after a dermatologic patient finds a dermatologist in a hospital to see a doctor, the dermatologist can know the condition of the patient, and then a medicine list is issued for the dermatologic patient. After the skin disease patient takes the medicine, the medicine is taken home to use and the treatment progress of the skin disease is observed, and no doctor participates in the subsequent process basically. The skin disease patients have no relevant knowledge or experience, and the whole course effect evaluation of the skin disease is difficult to carry out.
Disclosure of Invention
The invention aims to provide a quick, convenient and visual skin disease treatment effect automatic evaluation method and system.
The purpose of the invention is realized by the following technical scheme: an automatic evaluation method for the treatment effect of skin diseases comprises the following steps:
(1) collecting skin disease focus images of the same skin disease case at different times in the treatment process on time;
(2) processing and evaluating the acquired skin disease focus images each time by using a treatment effect evaluator to obtain treatment effect evaluation information of the skin disease case; the treatment effect evaluation information comprises treatment effect quantitative data of each focus image and a treatment effect time line (namely a treatment effect trend chart) of the skin disease case; the treatment effect quantification data comprises lesion area data of each lesion image and area data occupied by each grade color in each lesion image.
An automatic evaluation system for skin disease treatment effect comprises a client terminal and an automatic evaluation server for skin disease treatment effect;
the client terminal comprises a case establishing module, a data acquisition module, a storage module A and a data transmission module A;
the case establishing module is used for establishing a skin disease case;
the data acquisition module is used for acquiring skin disease focus images of the established skin disease cases;
the storage module A is used for storing the skin disease focus images acquired by the data acquisition module and storing treatment effect evaluation results;
the data transmission module A is used for transmitting the skin disease focus images acquired by the data acquisition module to the skin disease treatment effect automatic evaluation server and receiving a treatment effect evaluation result output by the skin disease treatment effect automatic evaluation server;
the automatic evaluation server for the skin disease treatment effect comprises a data transmission module B, a storage module B, an image preprocessing module, a treatment effect evaluator and a treatment effect evaluation result generation module;
the data transmission module B is used for receiving the dermatosis focus image transmitted by the client terminal, receiving the treatment effect evaluation result generated by the treatment effect evaluation result generation module and outputting the treatment effect evaluation result to the client terminal;
the storage module B is used for storing the skin disease focus images received by the data transmission module B and storing the treatment effect evaluation results generated by the treatment effect evaluation result generation module;
the image preprocessing module is used for removing redundant images in the storage module B through an image redundancy removing algorithm, and then rotating and normalizing the images after redundancy removing to be uniform in size;
the treatment effect evaluator is divided into a spatial dimension treatment effect evaluation module and a time dimension treatment effect evaluation module;
the spatial dimension treatment effect evaluation module obtains dermatosis treatment effect quantitative data of the dermatosis case by utilizing a treatment effect evaluation algorithm of spatial dimension;
the time dimension treatment effect evaluation module obtains a treatment effect time line of the skin disease case by using a time dimension treatment effect evaluation algorithm;
the treatment effect evaluation result generation module generates a treatment effect evaluation result according to the skin disease treatment effect quantitative data and the treatment effect timeline obtained by the treatment effect evaluator.
Compared with the prior art, the invention has the advantages that:
1. the automatic evaluation method for the treatment effect of the skin disease can enable a patient to quickly obtain accurate treatment effect evaluation information according to the focus images which are regularly collected and uploaded.
2. The automatic evaluation system for the treatment effect of the skin disease collects local images of affected parts of patients with the skin disease, then determines the treatment progress of the patients with the skin disease through the automatic evaluation server for the treatment effect of the skin disease based on deep learning, and gives related suggestions and cautions. The skin disease patient obtains accurate treatment information, can enhance the confidence of diagnosis and treatment schemes, eliminates uncertainty and improves the relationship between doctors and patients.
3. The automatic evaluation server for the skin disease treatment effect can complete the treatment effect evaluation of the skin disease case in the spatial dimension and the treatment effect evaluation in the time dimension.
4. The automatic evaluation server for the skin disease treatment effect completes the treatment effect evaluation of the spatial dimension, and can calculate the area ratio of the skin disease and the area ratio of the lesion classification (heavy, medium and light).
5. The skin disease treatment effect automatic evaluation server finishes the treatment effect evaluation of the time dimension, and can calculate the treatment effect time line and extract useful information of time.
6. The dermatosis patient can carry out threshold-free dermatosis whole-course effect evaluation on the premise of no medical knowledge, the patient can know more treatment conditions of the patient, and after the dermatosis patient knows detailed treatment information, anxiety can be relieved and the treatment scheme can be kept confident.
Drawings
FIG. 1 is a schematic flow chart of the automatic evaluation system for skin disease treatment effect of the present invention.
Fig. 2 is a step of the spatial dimension treatment effect evaluation module performing spatial dimension treatment effect evaluation.
Fig. 3 is a diagram of an image registration process of lesion images at different times of skin diseases in a spatial dimension treatment effect evaluation step.
Figure 4 is a flow diagram of the time dimension treatment effect assessment module performing time dimension treatment effect assessment,
fig. 5 is a timeline diagram of treatment effect.
Detailed Description
The invention is described in detail below with reference to the drawings and examples of the specification:
the first embodiment is as follows:
fig. 1-5 are schematic diagrams illustrating an embodiment of an automatic evaluation system for skin disease treatment effect according to the present invention.
The automatic evaluation system for the treatment effect of the skin disease provided by the invention collects the local image of the affected part of the skin disease patient, then determines the treatment progress of the skin disease patient through the automatic evaluation server for the treatment effect of the skin disease based on deep learning, and gives related suggestions and cautions. As shown in fig. 1, the flow principle of the automatic evaluation system for skin disease treatment effect is as follows:
(1) the user enters a certain established skin disease case in the client terminal (e.g. the whole skin disease assessment APP). (2) The method comprises the steps that the APP for the whole-process evaluation of the skin disease collects photos of symptoms of the skin disease, and the APP for the whole-process evaluation of the skin disease can automatically zoom and cut the photos to a preset size in the collection process. (3) And obtaining information related to the treatment effect of the skin disease by using a related skin disease treatment effect evaluation algorithm through the automatic evaluation server of the treatment effect of the skin disease. (4) The whole-course evaluation of skin diseases APP shows treatment effect evaluation information. The automatic evaluation system for skin disease treatment effect of the present invention is explained in more detail below:
the automatic evaluation system for the treatment effect of the skin disease comprises a client terminal and an automatic evaluation server for the treatment effect of the skin disease;
the client terminal comprises a case establishing module, a data acquisition module, an image detection module, a storage module A and a data transmission module A;
the case establishing module is used for establishing a skin disease case;
the data acquisition module is used for acquiring skin disease focus images of the established skin disease cases;
the image detection module is used for detecting whether the newly acquired skin disease focus image is matched with the currently operated skin disease case; the method for detecting the skin disease focus image by the image detection module comprises the following steps:
a. carrying out time correlation detection on the newly acquired focus image and the currently operated skin disease case, wherein the time correlation detection of the focus is preferably linear function or Gaussian function relation, and if the time correlation detection of the focus exceeds a threshold value, the time correlation detection fails, namely the newly acquired focus image does not belong to the focus image of the skin disease case;
b. carrying out focus correlation detection on a newly acquired focus image and a currently operated skin disease case, wherein the focus correlation detection firstly extracts edge features of the focus image; then, image registration is carried out on the focus edge features, if the registration feature matching rate is high (20 features are matched, the matching rate exceeds 20%), the focus correlation detection is successful, and if not, the detection is failed;
c. and (3) using a correlation discriminator to fuse and judge the time correlation and the focus correlation of the focus image sequence, wherein if the focus correlation (20%, 40%, 60%, 70% and 80%, the closer the time is, the higher the matching degree is) of the focus image sequence (such as a focus image sequence consisting of 5 focus images) conforms to the linear function or Gaussian function relationship of the time correlation, the final matching is successful, namely the newly acquired focus image is the focus image of the skin disease case.
The image detection module is set to prevent the image of a non-current case from being uploaded to the case, and when the image of tinea manus and pedis is uploaded in the eczema case, the image detection module automatically detects the wrong behavior and gives a wrong prompt.
The storage module A is used for storing the skin disease focus images acquired by the data acquisition module and storing treatment effect evaluation results;
the client terminal also comprises an image processing module, and the image processing module is used for automatically zooming or cutting the skin disease focus image acquired by the data acquisition module A to a preset size;
the data transmission module A is used for transmitting the skin disease focus image processed by the image processing module to the skin disease treatment effect automatic evaluation server and receiving a treatment effect evaluation result output by the skin disease treatment effect automatic evaluation server;
the client terminal may be an application installed on a mobile phone or a computer (i.e., APP, such as an APP for global skin disease assessment with the above-mentioned functions).
The automatic evaluation server for the skin disease treatment effect comprises a data transmission module B, a storage module B, an image preprocessing module, a treatment effect evaluator and a treatment effect evaluation result generation module;
the data transmission module B is used for receiving the dermatosis focus image transmitted by the client terminal, receiving the treatment effect evaluation result generated by the treatment effect evaluation result generation module and outputting the treatment effect evaluation result to the client terminal;
the storage module B is used for storing the skin disease focus images received by the data transmission module B and storing the treatment effect evaluation results generated by the treatment effect evaluation result generation module;
the image preprocessing module is used for removing redundant images in the storage module B through an image redundancy removing algorithm, and then rotating and normalizing the images after redundancy removing to be uniform in size;
the specific method for removing the redundant image in the image by the image redundancy removing algorithm may be as follows:
calculating the similarity SI between two adjacent images in a time sequence according to the following image redundancy removing algorithm;
wherein N is the image width, giPixel column, s, of source imageiThe image processing method comprises the steps of obtaining a target image pixel column, wherein an image with a front time sequence is a source image, and an image with a rear time sequence is a target image; the weighted average value SI' of the similarity SI between two images adjacent in time series is calculated according to the following formula:
SI′＝0.299SIr+0.587SIg+0.114SIb
wherein, SIrSI value of red color, SIgSI value of green, SIbSI value of blue;
and judging whether the weighted average value SI 'is in a preset similarity threshold range, if so, deleting the weighted average value SI', indicating that two adjacent images in the time sequence are similar, deleting the target image, if not, indicating that the two adjacent images in the time sequence are not similar, and keeping the two adjacent images in the time sequence.
The treatment effect evaluator is divided into a spatial dimension treatment effect evaluation module and a time dimension treatment effect evaluation module;
the spatial dimension treatment effect evaluation module utilizes a spatial dimension treatment effect evaluation algorithm to calculate and process the image preprocessed by the image preprocessing module so as to obtain dermatosis treatment effect quantitative data of dermatosis cases;
the treatment effect quantitative data comprises focus area data of each focus image and area data occupied by each grade color in each focus image;
the treatment effect evaluation algorithm for the spatial dimension comprises the following steps (as shown in fig. 2):
detecting the legality of a skin disease focus image: carrying out legality detection on the focus image and reserving the legal focus image, wherein the legal focus image requires that: the main body of the image is a picture of a focus of the skin disease, and the surface of the focus main body of the image does not contain covering objects, stains and water. If the detection does not meet the processing requirement, the error result is returned to the client terminal, and the client terminal performs error processing according to certain logic.
Preprocessing the images of the skin disease focus; the skin disease focus image preprocessing comprises the operations of enhancing image characteristics, enhancing image brightness, adjusting contrast and the like;
deleting non-focus main body in focus image;
image registration: adopting rotation and scaling operation to complete the registration operation of the skin lesion image and the historical lesion image of the case, and configuring all the lesion images under the case into the same spatial coordinate system;
the images must be registered or the evaluation of the effect of the space dimension of the lesion images cannot be carried out. The lesion image registration refers to the configuration of lesion images in the same spatial coordinate system. The randomly shot images generally have random factors such as angle rotation, focal length change and the like, and the front and rear focus images generally have obvious corresponding relation. The registration operation is to complete a simple registration operation of the lesion images by rotation and scaling operations in order to find a correspondence between the previous and subsequent lesion images, so that the angles of the lesion images and the sizes of the subjects in the images are the same (as shown in fig. 3).
Calculating the area of the focus: after the images are registered, respectively calculating the focus area of each focus image;
sixthly, calculating the area of the grading effect: according to the color grading of the skin diseases, calculating the area occupied by each grade color in each focus image, namely the grading effect area; and calculating the occupied area of each grade color in the focus image by adopting a point-by-point color fuzzy matching method. When the grading effect area is calculated, the focus of a focus image can be graded into a plurality of small focuses with different color depths according to the color of the skin disease, and then the area of each small focus is calculated. Wherein, the color grading can be divided into 5 grades according to the color degree of dark red, bright red, medium red, light red and reddish.
The time dimension treatment effect evaluation module utilizes a time dimension treatment effect evaluation algorithm to calculate and process the image preprocessed by the image preprocessing module so as to obtain a treatment effect time line of the skin disease case;
the time-dimension treatment effect evaluation algorithm comprises the following steps (as shown in fig. 4):
1) reading all focus images of the current skin disease case;
2) fusing core quantization indexes:
respectively calculating the total area of the focus edge (namely the focus area) of the focus image;
and II, respectively calculating the grading effect area of the focus image: classifying the focus of a focus image into a plurality of small focuses with different color depths according to the color of the skin disease, and then calculating the area of each small focus; wherein, the color grading can be divided into 5 grades according to the color degrees of dark red, bright red, medium red, light red and reddish, and the color weights of the 5 grades can be respectively set as 5 grades of weights, 4 grades of weights, 3 grades of weights, 2 grades of weights and 1 grade of weights;
fusing the focus area, the small focus area and the color weight into a core quantization index by using the following formula;
3) obtaining a treatment effect time line of the skin disease case corresponding to the quantitative index according to the core quantitative index (as shown in figure 5);
4) the subsequent treatment time is estimated.
The treatment effect evaluation result generation module generates a treatment effect evaluation result according to the skin disease treatment effect quantitative data and the treatment effect timeline obtained by the treatment effect evaluator.
The treatment effect evaluation result includes the following contents: the treatment effect timeline, the current state of the patient (namely the state of the latest uploaded dermatosis focus image, including the focus area data of the focus image and the area data occupied by each level color in the focus image), the corresponding medical advice, the patient attention and the like.
After the skin disease treatment effect automatic evaluation server finishes the treatment effect evaluation of the space dimension and the treatment effect evaluation of the time dimension, corresponding medical orders or cautions are given according to the treatment effect time line and the state of the skin disease patient. These data are presented after the APP receives them for the full-scale assessment of skin disease.
Example two:
an automatic evaluation method for the treatment effect of skin diseases comprises the following steps:
(1) collecting skin disease focus images of the same skin disease case at different times in the treatment process on time;
(2) processing and evaluating the acquired skin disease focus images each time by using a treatment effect evaluator to obtain treatment effect evaluation information of the skin disease case; the treatment effect evaluation information comprises treatment effect quantitative data of each focus image and a treatment effect time line (namely a treatment effect trend chart) of the skin disease case; the treatment effect quantification data comprises lesion area data of each lesion image and area data occupied by each grade color in each lesion image.
The treatment effect evaluator is used for carrying out time-dimension treatment effect evaluation and space-dimension treatment effect evaluation on the acquired skin disease focus images;
evaluating the treatment effect of the spatial dimension to obtain quantitative data of the treatment effect of the skin diseases;
the treatment effect assessment in the time dimension yields a treatment effect timeline.
The assessment of the treatment effect in the spatial dimension comprises the following steps (as shown in fig. 2):
preprocessing images of skin disease focuses; the skin disease focus image preprocessing comprises the operations of enhancing image characteristics, enhancing image brightness and contrast and the like;
deleting non-focus main body in focus image;
image registration: registering the skin disease focus image and the historical focus image of the case, and configuring all focus images of the case into the same spatial coordinate system;
fourthly, calculating the area of the focus: after the images are registered, respectively calculating the focus area of each focus image;
calculating the grading effect area: according to the color grading of the skin diseases, calculating the area occupied by each grade color in each focus image, namely the grading effect area; and calculating the occupied area of each grade color in the focus image by adopting a point-by-point color fuzzy matching method. When the grading effect area is calculated, the focus of a focus image can be graded into a plurality of small focuses with different color depths according to the color of the skin disease, and then the area of each small focus is calculated. Wherein, the color grading can be divided into 5 grades according to the color degree of dark red, bright red, medium red, light red and reddish.
The step of evaluating the treatment effect of the spatial dimension also comprises the detection of the legality of a focus image before the pretreatment of the skin disease focus image; detecting the legality of the focus image and keeping the legal focus image, wherein the legal focus image requires that: the main body of the image is a picture of a focus of the skin disease, and the surface of the focus main body of the image does not contain covering objects, stains and water.
And step three, finishing the registration of the focus image by adopting rotation and scaling operations.
The images must be registered or the evaluation of the effect of the space dimension of the lesion images cannot be carried out. The lesion image registration refers to the configuration of lesion images in the same spatial coordinate system. The randomly shot images generally have random factors such as angle rotation, focal length change and the like, and the front and rear focus images generally have obvious corresponding relation. The registration operation is to complete a simple registration operation of the lesion image by rotation and scaling operations in order to find the relationship between the front and rear lesion image objects, so that the angle of each lesion image and the size of the main body in the image are the same (as shown in fig. 3).
The evaluation of the treatment effect in the time dimension comprises the following steps (as shown in fig. 4):
1) reading all focus images of the current skin disease case;
2) fusing core quantization indexes:
respectively calculating the total area of the focus edge (namely the focus area) of the focus image;
and II, respectively calculating the grading effect area of the focus image: classifying the focus of a focus image into a plurality of small focuses with different color depths according to the color of the skin disease, and then calculating the area of each small focus; wherein, the color grading can be divided into 5 grades according to the color degrees of dark red, bright red, medium red, light red and reddish, and the color weights of the 5 grades can be respectively set as 5 grades of weights, 4 grades of weights, 3 grades of weights, 2 grades of weights and 1 grade of weights;
fusing the focus area, the small focus area and the color weight into a core quantization index by using the following formula;
3) obtaining a treatment effect time line of the skin disease case corresponding to the quantitative index according to the core quantitative index (as shown in figure 5);
4) the subsequent treatment time is estimated.
The automatic evaluation method for the skin disease treatment effect detects the skin disease focus image before processing and evaluating the collected skin disease focus image by using the treatment effect evaluator, and detects whether the newly collected skin disease focus image is matched with the currently operated skin disease case.
The method for detecting the dermatosis focus image comprises the following steps:
a. carrying out time correlation detection on the newly acquired focus image and the currently operated skin disease case, wherein the time correlation detection of the focus is preferably linear function or Gaussian function relation, and if the time correlation detection of the focus exceeds a threshold value, the time correlation detection fails, namely the newly acquired focus image does not belong to the focus image of the skin disease case;
b. carrying out focus correlation detection on a newly acquired focus image and a currently operated skin disease case, wherein the focus correlation detection firstly extracts edge features of the focus image; then, image registration is carried out on the focus edge features, if the registration feature matching rate is high (20 features are matched, the matching rate exceeds 20%), the focus correlation detection is successful, and if not, the detection is failed;
c. and (3) using a correlation discriminator to fuse and judge the time correlation and the focus correlation of the focus image sequence, wherein if the focus correlation (20%, 40%, 60%, 70% and 80%, the closer the time is, the higher the matching degree is) of the focus image sequence (such as a focus image sequence consisting of 5 focus images) conforms to the linear function or Gaussian function relationship of the time correlation, the final matching is successful, namely the newly acquired focus image is the focus image of the skin disease case.
It should be noted that the contents (such as image normalization, image automatic scaling or cropping, etc., total lesion edge area calculation of lesion image, etc.) not described in detail in this specification belong to the prior art known to those skilled in the art.
Claims (4)
1. An automatic evaluation method for the treatment effect of skin diseases, which is characterized in that: it comprises the following steps:
(1) collecting skin disease focus images of the same skin disease case at different times in the treatment process on time;
(2) processing and evaluating the acquired skin disease focus images each time by using a treatment effect evaluator to obtain treatment effect evaluation information of the skin disease case; the treatment effect evaluation information comprises treatment effect quantitative data of each focus image and a treatment effect timeline of the skin disease case; the treatment effect quantitative data comprises focus area data of each focus image and area data occupied by each grade color in each focus image;
the treatment effect evaluator performs treatment effect evaluation of a time dimension and treatment effect evaluation of a space dimension on the acquired skin disease focus image;
evaluating the treatment effect of the spatial dimension to obtain quantitative data of the treatment effect of the skin diseases;
evaluating the treatment effect in the time dimension to obtain a treatment effect timeline;
the assessment of the treatment effect in the spatial dimension comprises the following steps:
preprocessing images of skin disease focuses;
deleting non-focus main body in focus image;
image registration: registering the skin disease focus image and the historical focus image of the case, and configuring all focus images of the case into the same spatial coordinate system;
fourthly, calculating the area of the focus: after the images are registered, respectively calculating the focus area of each focus image;
calculating the grading effect area: according to the color grading of the skin diseases, calculating the area occupied by each grade color in each focus image, namely the grading effect area;
the assessment of the treatment effect in the time dimension comprises the following steps:
1) reading all focus images of the current skin disease case;
2) fusing core quantization indexes:
respectively calculating the total area of the focus edge of the focus image;
and II, respectively calculating the grading effect area of the focus image: classifying the focus of a focus image into a plurality of small focuses with different color depths according to the color of the skin disease, and then calculating the area of each small focus; wherein, the color grading can be divided into 5 grades according to the color degrees of dark red, bright red, medium red, light red and reddish, and the color weights of the 5 grades can be respectively set as 5 grades of weights, 4 grades of weights, 3 grades of weights, 2 grades of weights and 1 grade of weights;
fusing the focus area, the small focus area and the color weight into a core quantization index by using the following formula;
wherein n refers to the number of small lesions;
3) obtaining a treatment effect time line of the skin disease case corresponding to the quantitative index according to the core quantitative index;
4) estimating a subsequent treatment time;
before the treatment effect evaluator is used for processing and evaluating the collected skin lesion images, the skin lesion images are detected, and whether the newly collected skin lesion images are matched with the currently operated skin lesion examples is detected;
the method for detecting the dermatosis focus image comprises the following steps:
a. carrying out time correlation detection on the newly acquired focus image and the currently operated skin disease case, wherein the time correlation detection of the focus is a linear function or Gaussian function relation, and if the time correlation detection exceeds a threshold value, the time correlation detection fails;
b. carrying out focus correlation detection on a newly acquired focus image and a currently operated skin disease case, wherein the focus correlation detection firstly extracts edge features of the focus image; then, image registration is carried out on the focus edge features, if the registration feature matching rate is high, focus correlation detection is successful, and if not, detection is failed;
c. and (3) using a correlation discriminator to fuse and judge the time correlation and the focus correlation of the focus image sequence, and if the focus correlation of the focus image sequence conforms to the linear function or Gaussian function relationship of the time correlation, the final matching is successful, namely the newly acquired focus image is the focus image of the skin disease case.
2. The method for automatically evaluating the therapeutic effect of skin diseases according to claim 1, characterized in that: the step of evaluating the treatment effect of the spatial dimension also comprises the detection of the legality of a focus image before the pretreatment of the skin disease focus image; detecting the legality of the focus image and keeping the legal focus image, wherein the legal focus image requires that: the main body of the image is a picture of a focus of the skin disease, and the surface of the focus main body of the image does not contain covering objects, stains and water.
3. The method for automatically evaluating the therapeutic effect of skin diseases according to claim 1, characterized in that: and step three, finishing the registration of the focus image by adopting rotation and scaling operations.
4. An automatic evaluation system for skin disease treatment effect, characterized in that: the system comprises a client terminal and an automatic evaluation server for the treatment effect of the skin disease;
the client terminal comprises a case establishing module, a data acquisition module, a storage module A and a data transmission module A;
the case establishing module is used for establishing a skin disease case;
the data acquisition module is used for acquiring skin disease focus images of the established skin disease cases;
the storage module A is used for storing the skin disease focus images acquired by the data acquisition module and storing treatment effect evaluation results;
the data transmission module A is used for transmitting the skin disease focus image to the skin disease treatment effect automatic evaluation server and receiving a treatment effect evaluation result output by the skin disease treatment effect automatic evaluation server;
the automatic evaluation server for the skin disease treatment effect comprises a data transmission module B, a storage module B, an image preprocessing module, a treatment effect evaluator and a treatment effect evaluation result generation module;
the data transmission module B is used for receiving the dermatosis focus image transmitted by the client terminal, receiving the treatment effect evaluation result generated by the treatment effect evaluation result generation module and outputting the treatment effect evaluation result to the client terminal;
the storage module B is used for storing the skin disease focus images received by the data transmission module B and storing the treatment effect evaluation results generated by the treatment effect evaluation result generation module;
the image preprocessing module is used for removing redundant images in the storage module B through an image redundancy removing algorithm, and then rotating and normalizing the images after redundancy removing to be uniform in size;
the treatment effect evaluator is divided into a spatial dimension treatment effect evaluation module and a time dimension treatment effect evaluation module;
the spatial dimension treatment effect evaluation module utilizes a spatial dimension treatment effect evaluation algorithm to calculate and process the image preprocessed by the image preprocessing module so as to obtain dermatosis treatment effect quantitative data of dermatosis cases;
the time dimension treatment effect evaluation module utilizes a time dimension treatment effect evaluation algorithm to calculate and process the image preprocessed by the image preprocessing module so as to obtain a treatment effect time line of the skin disease case;
the treatment effect evaluation result generation module generates a treatment effect evaluation result according to the skin disease treatment effect quantitative data and the treatment effect timeline obtained by the treatment effect evaluator;
the treatment effect evaluation algorithm of the space dimension comprises the following steps:
detecting the legality of a skin disease focus image: carrying out legality detection on the focus image and reserving the legal focus image;
preprocessing the images of the skin disease focus;
deleting non-focus main body in focus image;
image registration: adopting rotation and scaling operation to complete the registration operation of the skin lesion image and the historical lesion image of the case, and configuring all the lesion images under the case into the same spatial coordinate system;
calculating the area of the focus: after the images are registered, respectively calculating the focus area of each focus image;
sixthly, calculating the area of the grading effect: according to the color grading of the skin diseases, calculating the area occupied by each grade color in each focus image, namely the grading effect area;
the time-dimension treatment effect evaluation algorithm comprises the following steps:
1) reading all focus images of the current skin disease case;
2) fusing core quantization indexes:
respectively calculating the total area of the focus edge of the focus image;
and II, respectively calculating the grading effect area of the focus image: classifying the focus of a focus image into a plurality of small focuses with different color depths according to the color of the skin disease, and then calculating the area of each small focus; wherein, the color grading can be divided into 5 grades according to the color degrees of dark red, bright red, medium red, light red and reddish, and the color weights of the 5 grades can be respectively set as 5 grades of weights, 4 grades of weights, 3 grades of weights, 2 grades of weights and 1 grade of weights;
fusing the focus area, the small focus area and the color weight into a core quantization index by using the following formula;
wherein n refers to the number of small lesions;
3) obtaining a treatment effect time line of the skin disease case corresponding to the quantitative index according to the core quantitative index;
4) estimating a subsequent treatment time;
the client terminal also comprises an image detection module; the image detection module is used for detecting whether the newly acquired skin disease focus image is matched with the currently operated skin disease case;
the method for detecting the skin disease focus image by the image detection module comprises the following steps:
a. carrying out time correlation detection on the newly acquired focus image and the currently operated skin disease case, wherein the time correlation detection of the focus is a linear function or Gaussian function relation, and if the time correlation detection exceeds a threshold value, the time correlation detection fails;
b. carrying out focus correlation detection on a newly acquired focus image and a currently operated skin disease case, wherein the focus correlation detection firstly extracts edge features of the focus image; then, image registration is carried out on the focus edge features, if the registration feature matching rate is high, focus correlation detection is successful, and if not, detection is failed;
c. and (3) using a correlation discriminator to fuse and judge the time correlation and the focus correlation of the focus image sequence, and if the focus correlation of the focus image sequence conforms to the linear function or Gaussian function relationship of the time correlation, the final matching is successful, namely the newly acquired focus image is the focus image of the skin disease case.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910813912.3A CN110544534B (en) | 2019-08-30 | 2019-08-30 | Method and system for automatically evaluating treatment effect of skin disease |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910813912.3A CN110544534B (en) | 2019-08-30 | 2019-08-30 | Method and system for automatically evaluating treatment effect of skin disease |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110544534A CN110544534A (en) | 2019-12-06 |
| CN110544534B true CN110544534B (en) | 2022-04-19 |
Family
ID=68712315
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910813912.3A Active CN110544534B (en) | 2019-08-30 | 2019-08-30 | Method and system for automatically evaluating treatment effect of skin disease |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110544534B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111430031B (en) * | 2020-05-18 | 2024-03-01 | 复旦大学附属中山医院 | Voxel and CT value multiplication integral algorithm system for evaluating lung cancer curative effect |
| CN112151136A (en) * | 2020-09-30 | 2020-12-29 | 上海依智医疗技术有限公司 | Medical data processing method and device and storage medium |
| CN113643227A (en) * | 2020-11-30 | 2021-11-12 | 马学召 | Skin care effect detection method and device based on deep learning technology |
| CN113298157A (en) * | 2021-05-28 | 2021-08-24 | 上海商汤智能科技有限公司 | Focus matching method and device, electronic equipment and storage medium |
| CN116269217B (en) * | 2023-02-10 | 2024-04-26 | 安徽医科大学 | Vitiligo treatment effect quantitative evaluation method based on deep learning |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201838001U (en) * | 2010-10-26 | 2011-05-18 | 冯睿 | Computer assisted image analysis and curative effect evaluation equipment for intervened low limb skin change |
| CN107049263A (en) * | 2017-06-14 | 2017-08-18 | 武汉理工大学 | Leucoderma condition-inference and cosmetic effect evaluating method and system based on image procossing |
| CN107292103A (en) * | 2017-06-19 | 2017-10-24 | 京东方科技集团股份有限公司 | A kind of prognostic chart picture generation method and device |
| CN108154503A (en) * | 2017-12-13 | 2018-06-12 | 西安交通大学医学院第附属医院 | A kind of leucoderma state of an illness diagnostic system based on image procossing |
| CN108648825A (en) * | 2018-05-30 | 2018-10-12 | 江苏大学附属医院 | A kind of leucoderma hickie appraisal procedure based on image recognition |
-
2019
- 2019-08-30 CN CN201910813912.3A patent/CN110544534B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201838001U (en) * | 2010-10-26 | 2011-05-18 | 冯睿 | Computer assisted image analysis and curative effect evaluation equipment for intervened low limb skin change |
| CN107049263A (en) * | 2017-06-14 | 2017-08-18 | 武汉理工大学 | Leucoderma condition-inference and cosmetic effect evaluating method and system based on image procossing |
| CN107292103A (en) * | 2017-06-19 | 2017-10-24 | 京东方科技集团股份有限公司 | A kind of prognostic chart picture generation method and device |
| CN108154503A (en) * | 2017-12-13 | 2018-06-12 | 西安交通大学医学院第附属医院 | A kind of leucoderma state of an illness diagnostic system based on image procossing |
| CN108648825A (en) * | 2018-05-30 | 2018-10-12 | 江苏大学附属医院 | A kind of leucoderma hickie appraisal procedure based on image recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110544534A (en) | 2019-12-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110544534B (en) | Method and system for automatically evaluating treatment effect of skin disease | |
| CN111417980B (en) | Three-dimensional medical image analysis method and system for identification of vertebral fractures | |
| CN107480677B (en) | Method and device for identifying interest region in three-dimensional CT image | |
| US20210098115A1 (en) | Aspect score estimating system and method | |
| Chanwimaluang et al. | Hybrid retinal image registration | |
| CN111028206A (en) | Prostate cancer automatic detection and classification system based on deep learning | |
| CN112102237A (en) | Brain tumor recognition model training method and device based on semi-supervised learning | |
| Shao et al. | Automated quality assessment of fundus images via analysis of illumination, naturalness and structure | |
| WO2009029638A1 (en) | Iris recognition | |
| Chlebus et al. | Deep learning based automatic liver tumor segmentation in CT with shape-based post-processing | |
| CN112613471B (en) | Face living body detection method, device and computer readable storage medium | |
| CN118172614B (en) | Ordered ankylosing spondylitis rating method based on supervised contrast learning | |
| CN112767374A (en) | Alzheimer disease focus region semantic segmentation algorithm based on MRI | |
| CN117408988A (en) | Artificial intelligence-based focus image analysis method and apparatus | |
| CN111797900A (en) | Arteriovenous classification method and device of OCT-A image | |
| CN118396990A (en) | Automatic analysis method, system, equipment and electronic medium for hospital radiology department images based on deep learning | |
| CN112529900B (en) | Method, device, terminal and storage medium for matching ROI in mammary gland image | |
| US20160058423A1 (en) | Apparatus and method for interpolating lesion detection | |
| CN113221700A (en) | Diagnosis similarity matching method and device based on medical examination report recognition | |
| CN111612749B (en) | Focus detection method and device based on lung image | |
| Matsopoulos et al. | Detection of glaucomatous change based on vessel shape analysis | |
| CN116703837B (en) | MRI image-based rotator cuff injury intelligent identification method and device | |
| Amit et al. | Automatic dual-view mass detection in full-field digital mammograms | |
| JP2020010804A (en) | Medical image diagnosis support apparatus, program, and medical image diagnosis support method | |
| Koukounis et al. | Retinal image registration based on multiscale products and optic disc detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |