Acne diagnosis and treatment prediction method and system, medium and computer program product

By combining multi-branch CNN and CatBoost models with patient information, we have achieved refined diagnosis and personalized treatment prediction for acne, solving the problem of relying on doctors' experience in existing technologies and improving the accuracy and efficiency of acne diagnosis and treatment.

CN122337543APending Publication Date: 2026-07-03ZHUHAI THIRD PEOPLES HOSPITAL (ZHUHAI CHRONIC DISEASE PREVENTION & TREATMENT CENT ZHUHAI OCCUPATIONAL DISEASE PREVENTION & TREATMENT HOSPITAL)
0 Cites 0 Cited by

Patent Information

Application Number
CN202511797538.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-02
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

In current technologies, the identification and diagnosis of acne mainly rely on doctors' subjective experience, lacking automated tools for refined management and personalized prediction of treatment plans, resulting in prolonged treatment cycles.

Method used

A multi-branch CNN architecture was used to extract facial image features from multiple perspectives. These features were then combined with patient information to classify the severity of acne. The CatBoost model was used to predict treatment outcomes, and the SHapley algorithm was used to interpret the treatment prediction results.

Benefits of technology

It achieved high accuracy in classifying acne severity (0.930 ± 0.109) and predicting treatment accuracy (0.833 ± 0.104), providing a reliable tool for personalized treatment planning and improving the accuracy and interpretability of treatment strategies.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an acne diagnosis and treatment prediction method and system, a medium and a computer program product, and relates to the technical field of artificial intelligence and medical health, and the key points of the technical scheme are as follows: obtaining multi-view face image data, extracting features through a multi-branch CNN architecture, extracting single-view features from each branch, and extracting single-view features from each branch; fusing the extracted single view features to obtain fused features, and inputting the fused features into a classification model to obtain an acne severity classification result; the age, gender and prescription information of a patient is obtained, the age, gender and prescription information of the patient and the acne severity classification result are input into the trained prediction model, and a treatment prediction result is obtained; and establishing a prediction model interpreter by adopting a SHapley algorithm, and interpreting a treatment prediction result. The framework provided by the invention fills up a gap between acne diagnosis and personalized treatment planning, and provides a reliable and explainable clinical tool for improving patient results and optimizing treatment strategies.
Need to check novelty before this filing date? Find Prior Art