CN114451870A - Pigment nevus malignant change risk monitoring system - Google Patents

Pigment nevus malignant change risk monitoring system Download PDF

Info

Publication number
CN114451870A
CN114451870A CN202210377495.4A CN202210377495A CN114451870A CN 114451870 A CN114451870 A CN 114451870A CN 202210377495 A CN202210377495 A CN 202210377495A CN 114451870 A CN114451870 A CN 114451870A
Authority
CN
China
Prior art keywords
real
time
monitored
nevus
pigmented nevus
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.)
Pending
Application number
CN202210377495.4A
Other languages
Chinese (zh)
Inventor
陈翔
黄凯
赵爽
李宜昕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xiangya Hospital of Central South University
Original Assignee
Xiangya Hospital of Central South University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xiangya Hospital of Central South University filed Critical Xiangya Hospital of Central South University
Priority to CN202210377495.4A priority Critical patent/CN114451870A/en
Publication of CN114451870A publication Critical patent/CN114451870A/en
Pending legal-status Critical Current

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
    • A61B5/444—Evaluating skin marks, e.g. mole, nevi, tumour, scar
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/107—Measuring physical dimensions, e.g. size of the entire body or parts thereof
    • A61B5/1072—Measuring physical dimensions, e.g. size of the entire body or parts thereof measuring distances on the body, e.g. measuring length, height or thickness
    • 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
    • A61B5/443—Evaluating skin constituents, e.g. elastin, melanin, water
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7271—Specific aspects of physiological measurement analysis
    • A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/74—Details of notification to user or communication with user or patient ; user input means
    • A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
    • 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
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • 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
    • G06T7/60—Analysis of geometric attributes
    • G06T7/64—Analysis of geometric attributes of convexity or concavity
    • 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
    • G06T7/68—Analysis of geometric attributes of symmetry
    • 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
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • 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
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • 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)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Theoretical Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • Surgery (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Molecular Biology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Artificial Intelligence (AREA)
  • Physiology (AREA)
  • Geometry (AREA)
  • Signal Processing (AREA)
  • Dermatology (AREA)
  • Psychiatry (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Quality & Reliability (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Evolutionary Computation (AREA)
  • Fuzzy Systems (AREA)
  • Mathematical Physics (AREA)
  • Measuring And Recording Apparatus For Diagnosis (AREA)

Abstract

The invention discloses a pigmented nevus malignant change risk monitoring system, which identifies and extracts a real-time skin damage image of a pigmented nevus to be monitored and a real-time boundary contour thereof from a real-time picture of the pigmented nevus to be monitored; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the extracted real-time boundary contour; calculating the real-time color distribution uniformity and real-time actual size value of the pigmented nevus to be monitored in the real-time skin damage image, acquiring the actual size value of the pigmented nevus to be monitored in the past preset time period, and calculating the real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored in the past preset time period; and judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored. The method can accurately predict the risk of converting the pigmented nevus into the melanoma.

Description

Pigment nevus malignant change risk monitoring system
Technical Field
The invention relates to the field of intelligent monitoring of pigmented nevus, in particular to a system for monitoring the malignant change risk of pigmented nevus.
Background
Pigmented nevus, a benign tumor formed by locally increasing and gathering nevus cells, is common and is seen by almost everyone. The pigmented nevus has various shapes, colors and sizes, and may or may not be accompanied by hair growth. Congenital pigmented nevus has large or small size, and can be divided into boundary nevus, intradermal nevus and mixed nevus according to the position of nevus cells on the skin. The acquired pigmented nevus is basically expressed as macula or pimple with diameter less than 6mm, with clear boundary, regular edge, uniform color, and brown, black, etc. There are few accompanying symptoms such as itching and pain.
The location and characteristics of the mole considerably determine the magnitude of the risk of developing a malignancy. For Asians, pigmented nevi at the extremities of limbs are prone to malignant changes due to long-term friction; pigmented nevus (onychomycosis) growing under the nail also has a high risk of malignant change, and needs to be followed up closely and treated as soon as possible; pigmented nevi on exposed parts of the head, face, limbs, etc. may be converted into malignant melanoma if they are exposed to factors such as sunlight for a long time. In addition, "nevi" are stable for long periods without abnormalities, generally with a high likelihood of being benign; if the disease rapidly increases or ulcerates in a short period of time, pruritus and pain occur, or peripheral satellite foci (lentigo) occur, the possibility of malignant change is often indicated.
To solve the above problems, researchers train artificial intelligence machines to recognize skin cancer by showing 100,000 more images of malignant melanoma and benign nevi to the artificial intelligence machines. And the performance of the intelligent machine was compared to that of 58 international dermatologists, who could only detect on average 86.6% melanoma accurately when the intelligent machine was adjusted to the same level as the doctor to correctly identify benign nevi (71.3%). The intelligent diagnosis is found to reduce the missed diagnosis rate of melanoma and greatly reduce the misdiagnosis rate of benign nevus, so that the imaging difference of benign pigmented nevus and malignant melanoma and the possibility of machine identification are proved.
Although the existing intelligent diagnosis method can accurately distinguish melanoma and pigmented nevus, the risk prediction of benign-to-malignant transformation of pigmented nevus is lacked; and the early identification of the pigmented nevus with high malignant risk can improve the cure success rate of the pigmented nevus, and has important significance for the prognosis of patients.
Therefore, how to accurately monitor the risk of nevus pigmentosus degeneration has become a technical problem to be urgently solved by those skilled in the art.
Disclosure of Invention
The invention provides a pigmented nevus malignant change risk monitoring system, which is used for solving the technical problem that the prior art is lack of risk prediction on the benign to malignant change of pigmented nevus.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
a pigmented nevus malignant risk monitoring system, comprising:
an acquisition module: the system is used for acquiring a real-time picture of the pigmented nevus to be monitored;
a contour monitoring module: the real-time image processing system is used for identifying and extracting a real-time skin damage image of the pigmented nevus to be monitored and a real-time boundary contour of the pigmented nevus from the real-time image; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the real-time boundary contour;
a color monitoring module: the real-time color classification and the distribution area of each color of the pigmented nevus to be monitored in the real-time skin damage image are identified, and the real-time color distribution uniformity of the pigmented nevus to be monitored is determined according to the real-time color classification and the distribution area of each color;
size monitoring module: the system is used for calculating a real-time actual size value of the pigmented nevus to be monitored according to the real-time diameter, acquiring a past preset time period actual size value of the pigmented nevus to be monitored, and calculating a real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored and the past preset time period actual size value of the pigmented nevus to be monitored;
a prediction module: the method is used for judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored.
Preferably, the contour monitoring module comprises a feature recognition model and a contour segmentation model;
the characteristic identification model is a deep learning neural network model, is obtained by training a plurality of clinical images and skin mirror images marked with pigmented nevus and skin damage areas thereof, and is used for marking the pigmented nevus to be monitored and the real-time skin damage areas thereof on a real-time picture of the pigmented nevus to be monitored;
the contour segmentation model is constructed based on an image segmentation algorithm and is used for identifying the boundary contour of the real-time skin damage area from a real-time picture marked with the pigmented nevus to be monitored and the real-time skin damage area thereof and segmenting the real-time skin damage area according to the boundary contour.
Preferably, the contour monitoring module includes a symmetry degree calculation module, and the symmetry degree calculation module is configured to perform mirror image processing on the real-time boundary contour to obtain a mirror image of the real-time boundary contour; and aligning the centers of the real-time boundary contour and the mirror image thereof, and calculating the contact ratio of the real-time boundary contour and the mirror image thereof as the real-time symmetry of the pigmented nevus to be monitored.
Preferably, the contour monitoring module further includes a boundary smoothness calculation module, where the boundary smoothness calculation module is configured to perform equidistant sampling on the real-time boundary contour, connect each sampling point and its adjacent sampling points to form multiple tangent lines, calculate a slope of each tangent line, calculate a tangent included angle of each tangent line and its adjacent tangent line according to the slope of each tangent line and its adjacent tangent line, divide each tangent included angle by 180 degrees, perform normalization, and calculate a variance of all tangent included angles as the real-time boundary smoothness of the nevus pigment to be monitored.
Preferably, the color distribution uniformity is a gray uniformity, and the contour color monitoring module is configured to convert the pigmented nevus to be monitored into a gray image, and calculate the gray uniformity of the gray image by using an LBP method.
Preferably, the prediction module is configured to count the number of terms of the pigmented nevus to be monitored, which satisfy the following characteristic values of the risk of malignant change, and determine the size of the risk of malignant change of the pigmented nevus to be monitored according to the number of terms of the characteristic values of the risk of malignant change that satisfy:
firstly, the real-time diameter is larger than a preset diameter change threshold value;
secondly, the real-time symmetry degree is smaller than a preset symmetrical aversion threshold value;
thirdly, the real-time boundary smoothness is smaller than a preset smooth aversion threshold value;
fourthly, the real-time color distribution uniformity is smaller than a preset color uniformity aversion threshold value;
fifthly, the real-time growth rate is larger than a preset growth rate change threshold.
Preferably, the prediction module judges the nevus pigment malignant change risk to be monitored according to the number of terms of the satisfied malignant change risk characteristic value, and the method is implemented according to the following rules:
when the number of terms of the satisfied malignant change risk characteristic values is 5, judging that the malignant change risk of the pigmented nevus to be monitored is high risk;
when the number of terms of the satisfied malignant change risk characteristic values is 3 or 4, judging that the nevus pigment malignant change risk to be monitored is an intermediate risk;
when the number of terms of the satisfied malignant change risk characteristic values is 2, judging that the malignant change risk of the pigmented nevus to be monitored is low risk;
and when the number of terms of the satisfied malignant change risk characteristic values is less than 1 term, judging that the malignant change risk of the pigmented nevus to be monitored is extremely low risk.
Preferably, the prediction module is configured to obtain epidemiological risk factors of a patient corresponding to the pigmented nevus to be monitored, and comprehensively determine whether the pigmented nevus to be monitored has a malignant transformation risk by combining the growth rate of the pigmented nevus to be monitored, and the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity, and the real-time color depth of the pigmented nevus to be monitored.
The invention has the following beneficial effects:
1. the pigmented nevus malignant change risk monitoring system identifies and extracts a real-time skin damage image of the pigmented nevus to be monitored and a real-time boundary contour thereof from a real-time picture of the pigmented nevus to be monitored; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the extracted real-time boundary contour; calculating the real-time color distribution uniformity and real-time actual size value of the pigmented nevus to be monitored in the real-time skin damage image, acquiring the actual size value of the pigmented nevus to be monitored in the past preset time period, and calculating the real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored and the actual size value of the past preset time period; and judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored. Compared with the prior art, the method can accurately predict the risk of the pigmented nevus to be converted into the melanoma, is beneficial to preventing the pigmented nevus to be converted into the melanoma, and reduces the management cost of the skin tumor.
In a preferable scheme, the method can further improve the accuracy of the risk prediction of the pigmented nevus to melanoma transition by analyzing the conformity degree of the long-term monitoring image of the pigmented nevus of the patient and the ABCDE principle and combining with the epidemiological risk factors of the patient to evaluate the possibility of the pigmented nevus malignant change.
In addition to the objects, features and advantages described above, other objects, features and advantages of the present invention are also provided. The present invention will be described in further detail below with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of the working process of the nevus pigment malignant risk monitoring system in the preferred embodiment of the present invention.
Detailed Description
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
The first embodiment is as follows:
as shown in fig. 1, this embodiment discloses a pigmented nevus malignant risk monitoring system, which includes:
an acquisition module: the system is used for acquiring a real-time picture of the pigmented nevus to be monitored;
a contour monitoring module: the real-time image processing system is used for identifying and extracting a real-time skin damage image of the pigmented nevus to be monitored and a real-time boundary contour of the pigmented nevus from the real-time image; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the real-time boundary contour;
a color monitoring module: the real-time color classification and the distribution area of each color of the pigmented nevus to be monitored in the real-time skin damage image are identified, and the real-time color distribution uniformity of the pigmented nevus to be monitored is determined according to the real-time color classification and the distribution area of each color;
size monitoring module: the system is used for calculating a real-time actual size value of the pigmented nevus to be monitored according to the real-time diameter, acquiring a past preset time period actual size value of the pigmented nevus to be monitored, and calculating a real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored and the past preset time period actual size value of the pigmented nevus to be monitored;
a prediction module: the method is used for judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored.
The pigmented nevus malignant change risk monitoring system identifies and extracts a real-time skin damage image of the pigmented nevus to be monitored and a real-time boundary contour thereof from a real-time picture of the pigmented nevus to be monitored; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the extracted real-time boundary contour; calculating the real-time color distribution uniformity and real-time actual size value of the pigmented nevus to be monitored in the real-time skin damage image, acquiring the actual size value of the pigmented nevus to be monitored in the past preset time period, and calculating the real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored and the actual size value of the past preset time period; and judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored. Compared with the prior art, the method can accurately predict the risk of the pigmented nevus to be converted into the melanoma, is beneficial to preventing the pigmented nevus to be converted into the melanoma, and reduces the management cost of the skin tumor.
The second embodiment:
the second embodiment is a preferred embodiment of the first embodiment, and is different from the first embodiment in that specific steps of the pigmented nevus malignant change risk monitoring system are introduced, and the method includes:
the invention aims to provide a pigmented nevus malignant change risk monitoring system based on multi-mode data, which can evaluate the possibility of pigmented nevus malignant change by analyzing the conformity degree of long-term monitoring images of the pigmented nevus of a patient and an ABCDE principle and combining the medical history information of the patient. The method makes up the defects, realizes the intelligent prediction of the risk of the pigmented nevus to the melanoma, guides the early intervention of patients and doctors, reduces the management cost of skin tumors, and reduces the risks of enlarged operation range, poorer prognosis, possible involvement of multiple organs and even death caused by the malignant change of the pigmented nevus. The method is convenient and fast, has high accuracy, and can monitor the growth condition of the pigmented nevus for a long time at any time, prevent the transformation of the pigmented nevus to the melanoma, and reduce the missed diagnosis rate of the melanoma.
The ABCDE principle is a method for preliminarily judging the malignant change risk of the pigmented nevus by the visual characteristics of skin lesions. A (asymmetry) represents symmetry, pigmented nevi are generally symmetrical; b (Border) represents the boundary, the mole is a general boundary rule; c (color) represents the color, and the pigmented nevus is generally single in color; d (diameter) represents diameter, nevus generally <6mm in diameter, E (Evalving) represents lesion growth, and nevus generally does not grow progressively. In addition, sex, age, phenotypic susceptibility (large nevus count), personal medical history, concurrent disease, genetic susceptibility, environmental factors are all recognized melanoma risk factors in the south.
In this embodiment, the system for monitoring the risk of nevus pigmentosus malignant transformation specifically includes the following steps:
the first step is as follows: data collection
Collect clinical image of pigmented nevus and melanoma respectively and each 1 ten thousand of dermatoscope images that correspond and regard as the training set for the machine carries out the degree of depth study to the image characteristic of pigmented nevus and melanoma, carries out the label to every picture, marks out the skin damage region, supplies the machine to carry out feature recognition.
Second step, risk evaluation standard and model are formulated
(1) Selecting a model suitable for the characteristic recognition of the pigmented nevus and melanoma from the existing models for the characteristic analysis of the pigmented nevus, firstly, accurately extracting a skin damage image from a clinical image and a skin mirror image and positioning the boundary of the skin damage image, and analyzing the symmetry, the boundary, the color and the diameter of the skin damage by a machine according to an ABCDE principle. The analysis method is as follows:
symmetry: calculating the contact ratio of the skin damage image after horizontally turning 180 degrees with the original skin damage boundary, wherein the higher the contact ratio is, the stronger the symmetry is, and the lower the malignant change possibility of the pigmented nevus is;
boundary: extracting a skin damage boundary, measuring the smoothness of a boundary curve, and determining that the smoother the boundary, the clearer the nevus pigment malignant change probability is; early manifestations of melanoma are often incomplete or irregular borders, such as small corners, or small defects, etc. And the pigmented nevus is basically not incomplete at the boundary.
Color: the color of the skin damage is extracted for analysis, and a plurality of obviously different colors can be extracted to prove that the possibility of the nevus pigment malignant change is increased; the pigmented nevi are all substantially uniform in color, but malignant melanoma often has a less uniform black color, and may be a variegated color, either black, dark black, light black, or even red, blue, white ink.
(2) Design an intelligent monitoring model for detect the long-term condition of growing of pigmented nevus, the patient needs the fixed distance of shooing through the image acquisition of this system regular collection pigmented nevus, every turn image acquisition to monitor the condition of growing of pigmented nevus better, if the pigmented nevus is the trend that obviously grows up in short-term, then suggestion this malignant change possibility of pigmented nevus increases. In addition, the patient can input epidemiological risk factors of the pigmented nevus malignant change through the monitoring system, so that the risk of the pigmented nevus malignant change is comprehensively analyzed.
(3) The two models are combined to form the system for monitoring and early warning the nevus pigmentosus malignant change risk, and the early warning standard is as follows:
coefficient of risk "ABCDE principle" for prompting malignant change (item) requirement Epidemiological risk factors suggest a malignant change requirement
High risk In accordance with 5 According to 3 items or more
Middle risk In accordance with 3 to 4 In accordance with item 2 and below
Low risk In accordance with 2 In accordance with item 2 and below
Extremely low risk In accordance with item 1 and below In accordance with item 2 and below
Wherein, the ABCDE principle prompts that the malignant change (item) is:
1) the real-time diameter is larger than a preset diameter change threshold value; 2) the real-time symmetry degree is smaller than a preset symmetric aversion threshold; 3) the real-time boundary smoothness is smaller than a preset smooth misconceptive threshold; 4) the real-time color distribution uniformity is smaller than a preset color uniformity change threshold value; 5) the real-time growth rate is greater than a preset growth rate change threshold;
wherein, epidemiological risk factors suggest that the malignant change (item) is:
1) skin color is light; 2) long-term ultraviolet irradiation or sunburn; 3) abnormal number or size of nevi; 4) immunosuppression; (5) family history of skin tumors; 6) history of previous skin tumors; 7) the medical history is more than half a year.
For any risk coefficient A, when the patient simultaneously meets the corresponding 'ABCDE principle' prompt aversion requirement and the epidemiological risk factor prompts aversion, judging that the patient has the risk coefficient, and if the patient A simultaneously meets 5 items in the 'ABCDE principle' prompt aversion and the epidemiological risk factor prompts aversion more than 3 items, judging that the patient has high risk.
Third step model validation
After the model is trained by the training set, a batch of pigmented nevus clinical images and melanoma clinical images which are not identified and learned by a machine are arranged as a verification set, clinical application is simulated, a patient and a system interact, and the system gives a risk early warning result and compares the risk early warning result with an actual diagnosis result. The evaluation indexes are as follows: the skin lesion image of a melanoma patient diagnosed can be evaluated as high risk, the pigmented nevus image which is recommended to be excised by a doctor is evaluated as medium risk/high risk, and the common pigmented nevus image is evaluated as low risk/extremely low risk.
Compared with the prior art, the invention has the following advantages:
1. the ABCDE principle for evaluating the malignant change risk of the pigmented nevus is introduced, the risk of the malignant change of the pigmented nevus into melanoma is predicted by analyzing different indexes of a pigmented nevus image, and early warning signals are given to doctors and patients in time, so that early discovery, early diagnosis and early intervention of skin malignant tumors are facilitated.
2. A long-term detection system is introduced, the growth condition of the pigmented nevus is monitored for a long term, evaluation is carried out through the dynamic process of the skin lesion, and the method is not limited to static skin lesion images. In addition, the invention records the epidemiological information of the patient, is not limited to the single-mode data of the image, has more comprehensive risk evaluation standard and more accurate and credible early warning result.
3. The artificial intelligence is applied to the pigmented nevus malignant change risk prediction decision for the first time, so that the anxiety of the patient on benign pigmented nevus can be relieved, early intervention on melanoma is realized, the management cost of melanoma is reduced, the prognosis result of the patient is optimized, and the skin health level of people is improved.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (8)

1. A pigmented nevus malignant risk monitoring system, characterized by comprising:
an acquisition module: the system is used for acquiring a real-time picture of the pigmented nevus to be monitored;
a contour monitoring module: the real-time image processing system is used for identifying and extracting a real-time skin damage image of the pigmented nevus to be monitored and a real-time boundary contour of the pigmented nevus from the real-time image; calculating the real-time diameter, the real-time symmetry and the real-time boundary smoothness of the pigmented nevus to be monitored according to the real-time boundary contour;
a color monitoring module: the real-time color classification and the distribution area of each color of the pigmented nevus to be monitored in the real-time skin damage image are identified, and the real-time color distribution uniformity of the pigmented nevus to be monitored is determined according to the real-time color classification and the distribution area of each color;
size monitoring module: the system is used for calculating a real-time actual size value of the pigmented nevus to be monitored according to the real-time diameter, acquiring a past preset time period actual size value of the pigmented nevus to be monitored, and calculating a real-time growth rate of the pigmented nevus to be monitored according to the real-time actual size value of the pigmented nevus to be monitored and the past preset time period actual size value of the pigmented nevus to be monitored;
a prediction module: the method is used for judging whether the pigmented nevus to be monitored has a malignant change risk or not according to the real-time diameter, the real-time symmetry, the real-time boundary smoothness, the real-time color distribution uniformity and the real-time growth rate of the pigmented nevus to be monitored.
2. The nevus pigmentosus malignant risk monitoring system according to claim 1, wherein the contour monitoring module comprises a feature recognition model and a contour segmentation model;
the characteristic identification model is a deep learning neural network model, is obtained by training a plurality of clinical images and skin mirror images marked with pigmented nevus and skin damage areas thereof, and is used for marking the pigmented nevus to be monitored and the real-time skin damage areas thereof on a real-time picture of the pigmented nevus to be monitored;
the contour segmentation model is constructed based on an image segmentation algorithm and is used for identifying the boundary contour of the real-time skin damage area from a real-time picture marked with the pigmented nevus to be monitored and the real-time skin damage area thereof and segmenting the real-time skin damage area according to the boundary contour.
3. The system according to claim 1, wherein the contour monitoring module comprises a symmetry degree calculation module, and the symmetry degree calculation module is configured to perform mirror image processing on the real-time boundary contour to obtain a mirror image of the real-time boundary contour; and aligning the centers of the real-time boundary contour and the mirror image thereof, and calculating the contact ratio of the real-time boundary contour and the mirror image thereof as the real-time symmetry of the pigmented nevus to be monitored.
4. The nevus pigmentosus malignant change risk monitoring system according to claim 3, wherein the contour monitoring module further comprises a boundary smoothness calculation module, the boundary smoothness calculation module is configured to sample the real-time boundary contour at equal intervals, connect each sampling point and its adjacent sampling points to form a plurality of tangents, calculate a slope of each tangent, calculate a tangent angle between each tangent and its adjacent tangent according to the slope of each tangent and its adjacent tangent, divide each tangent angle by 180 degrees, perform normalization, and calculate a variance of all tangent angles as the real-time boundary smoothness of the nevus pigmentosus to be monitored.
5. The nevus pigmentosus malignant risk monitoring system according to claim 3, wherein the color distribution uniformity is a gray scale uniformity, and the contour color monitoring module is configured to convert the nevus pigmentosus to be monitored into a gray scale image and calculate the gray scale uniformity of the gray scale image by using an LBP method.
6. The nevus pigmentosus malignant change risk monitoring system according to claim 1, wherein the prediction module is configured to count the number of terms of the nevus pigmentosus to be monitored that satisfy the following malignant change risk characteristic values, and determine the magnitude of the nevus pigmentosus malignant change risk according to the number of terms of the satisfied malignant change risk characteristic values:
firstly, the real-time diameter is larger than a preset diameter change threshold value;
secondly, the real-time symmetry degree is smaller than a preset symmetrical aversion threshold value;
thirdly, the real-time boundary smoothness is smaller than a preset smooth aversion threshold value;
fourthly, the real-time color distribution uniformity is smaller than a preset color uniformity aversion threshold value;
fifthly, the real-time growth rate is larger than a preset growth rate change threshold.
7. The nevus pigmentosus malignant risk monitoring system according to claim 6, wherein the prediction module determines the nevus pigmentosus malignant risk to be monitored according to the number of terms of the satisfied malignant risk characteristic values, and the determination is implemented by the following rules:
when the number of terms of the satisfied malignant change risk characteristic values is 5, judging that the malignant change risk of the pigmented nevus to be monitored is high risk;
when the number of terms of the satisfied malignant change risk characteristic values is 3 or 4, judging that the malignant change risk of the pigmented nevus to be monitored is a middle risk;
when the number of terms of the satisfied malignant change risk characteristic values is 2, judging that the malignant change risk of the pigmented nevus to be monitored is low risk;
and when the number of terms of the satisfied malignant change risk characteristic values is less than 1 term, judging that the malignant change risk of the pigmented nevus to be monitored is extremely low risk.
8. The system according to claim 5, wherein the prediction module is configured to obtain epidemiological risk factors of the mole to be monitored corresponding to the patient, and comprehensively determine whether there is a risk of malignant change in the mole to be monitored by combining the growth rate of the mole to be monitored, and the real-time diameter, real-time symmetry, real-time boundary smoothness, and real-time color distribution uniformity of the mole to be monitored.
CN202210377495.4A 2022-04-12 2022-04-12 Pigment nevus malignant change risk monitoring system Pending CN114451870A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210377495.4A CN114451870A (en) 2022-04-12 2022-04-12 Pigment nevus malignant change risk monitoring system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210377495.4A CN114451870A (en) 2022-04-12 2022-04-12 Pigment nevus malignant change risk monitoring system

Publications (1)

Publication Number Publication Date
CN114451870A true CN114451870A (en) 2022-05-10

Family

ID=81417848

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210377495.4A Pending CN114451870A (en) 2022-04-12 2022-04-12 Pigment nevus malignant change risk monitoring system

Country Status (1)

Country Link
CN (1) CN114451870A (en)

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101344924A (en) * 2007-07-12 2009-01-14 株式会社理光 Image processing apparatus, image processing method, and computer program product
CN101756696A (en) * 2009-12-31 2010-06-30 中国人民解放军空军总医院 Multiphoton skin lens image automatic analytical system and method for diagnosing malignant melanoma by using same system
WO2011087807A2 (en) * 2009-12-22 2011-07-21 Health Discovery Corporation System and method for remote melanoma screening
US20120008838A1 (en) * 2000-08-07 2012-01-12 Health Discovery Corporation System and method for remote melanoma screening
US20140126787A1 (en) * 2012-11-08 2014-05-08 Lasarow Healthcare Technologies Limited Device and method for generating melanoma risk assessments
US20170231550A1 (en) * 2014-08-25 2017-08-17 Singapore University Of Technology And Design Method and device for analysing an image
CN107909566A (en) * 2017-10-28 2018-04-13 杭州电子科技大学 A kind of image-recognizing method of the cutaneum carcinoma melanoma based on deep learning
WO2018122793A1 (en) * 2016-12-29 2018-07-05 Universita' Degli Studi Di Padova Method and device for a three-dimensional mapping of a patient's skin for supporting the melanoma diagnosis
CN108510502A (en) * 2018-03-08 2018-09-07 华南理工大学 Melanoma picture tissue segmentation methods based on deep neural network and system
CN110399902A (en) * 2019-06-27 2019-11-01 华南师范大学 A kind of method of melanoma texture feature extraction
CN112365485A (en) * 2020-11-19 2021-02-12 同济大学 Melanoma identification method based on Circular LBP and color space conversion algorithm
CN112950569A (en) * 2021-02-25 2021-06-11 平安科技（深圳）有限公司 Melanoma image recognition method and device, computer equipment and storage medium
CN113487568A (en) * 2021-07-05 2021-10-08 陕西科技大学 Liver surface smoothness measuring method based on differential curvature

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120008838A1 (en) * 2000-08-07 2012-01-12 Health Discovery Corporation System and method for remote melanoma screening
CN101344924A (en) * 2007-07-12 2009-01-14 株式会社理光 Image processing apparatus, image processing method, and computer program product
WO2011087807A2 (en) * 2009-12-22 2011-07-21 Health Discovery Corporation System and method for remote melanoma screening
CN101756696A (en) * 2009-12-31 2010-06-30 中国人民解放军空军总医院 Multiphoton skin lens image automatic analytical system and method for diagnosing malignant melanoma by using same system
US20140126787A1 (en) * 2012-11-08 2014-05-08 Lasarow Healthcare Technologies Limited Device and method for generating melanoma risk assessments
US20170231550A1 (en) * 2014-08-25 2017-08-17 Singapore University Of Technology And Design Method and device for analysing an image
WO2018122793A1 (en) * 2016-12-29 2018-07-05 Universita' Degli Studi Di Padova Method and device for a three-dimensional mapping of a patient's skin for supporting the melanoma diagnosis
CN107909566A (en) * 2017-10-28 2018-04-13 杭州电子科技大学 A kind of image-recognizing method of the cutaneum carcinoma melanoma based on deep learning
CN108510502A (en) * 2018-03-08 2018-09-07 华南理工大学 Melanoma picture tissue segmentation methods based on deep neural network and system
CN110399902A (en) * 2019-06-27 2019-11-01 华南师范大学 A kind of method of melanoma texture feature extraction
CN112365485A (en) * 2020-11-19 2021-02-12 同济大学 Melanoma identification method based on Circular LBP and color space conversion algorithm
CN112950569A (en) * 2021-02-25 2021-06-11 平安科技（深圳）有限公司 Melanoma image recognition method and device, computer equipment and storage medium
CN113487568A (en) * 2021-07-05 2021-10-08 陕西科技大学 Liver surface smoothness measuring method based on differential curvature

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
黄云超，等: "《临床肿瘤内科学（第二版）》", 31 May 2017, 云南科技出版社 *

Similar Documents

Publication Publication Date Title
CN110097034B (en) Intelligent face health degree identification and evaluation method
Khan et al. Classification of melanoma and nevus in digital images for diagnosis of skin cancer
US7689016B2 (en) Automatic detection of critical dermoscopy features for malignant melanoma diagnosis
Smaoui et al. A developed system for melanoma diagnosis
CN101756696B (en) Multiphoton skin lens image automatic analytical system and method for diagnosing malignant melanoma by using same system
US11875479B2 (en) Fusion of deep learning and handcrafted techniques in dermoscopy image analysis
Jaworek-Korjakowska Computer‐aided diagnosis of micro‐malignant melanoma lesions applying support vector machines
Kurugol et al. Automated delineation of dermal–epidermal junction in reflectance confocal microscopy image stacks of human skin
CN112017743B (en) Automatic generation platform and application of disease risk evaluation report
Xu et al. Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients
KR20090010555A (en) Computer aided diagnosis system for malignant melanoma using image analysis
Mahmouei et al. Biologically inspired QuadTree color detection in dermoscopy images of melanoma
Vocaturo et al. Features for melanoma lesions characterization in computer vision systems
Madooei et al. Incorporating Colour Information for Computer‐Aided Diagnosis of Melanoma from Dermoscopy Images: A Retrospective Survey and Critical Analysis
Singh et al. Designing a retrieval-based diagnostic aid using effective features to classify skin lesion in dermoscopic images
CN104933723B (en) Tongue image dividing method based on rarefaction representation
Saxena et al. Study of Computerized Segmentation & Classification Techniques: An Application to Histopathological Imagery
KR20180045473A (en) System, method and computer program for melanoma detection using image analysis
Rastghalam et al. Skin melanoma detection in microscopic images using HMM-based asymmetric analysis and expectation maximization
Dey et al. Red-plane asymmetry analysis of breast thermograms for cancer detection
Dehmeshki et al. Classification of disease subgroup and correlation with disease severity using magnetic resonance imaging whole-brain histograms: application to magnetization transfer ratios and multiple sclerosis
CN117808786A (en) Retinal artery branch angle change correlation prediction method
Taher et al. Identification of lung cancer based on shape and color
Dika et al. Advantages of manual and automatic computer-aided compared to traditional histopathological diagnosis of melanoma: a pilot study
CN114451870A (en) Pigment nevus malignant change risk monitoring system

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20220510

RJ01 Rejection of invention patent application after publication