AI Myopia Regression Prediction Using Fundus Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The long-term effectiveness and side effects of vision correction surgeries, such as myopia regression, are unpredictable, leading to significant human, social, and economic costs due to the difficulty in diagnosing complications without long-term observation.
Innovation Solution
An electronic device utilizing an artificial intelligence machine learning model constructed from preoperative data and fundus photography to predict the possibility of myopia regression after vision correction surgery, by collecting and processing target data to determine the likelihood of regression based on input from multiple machine learning models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If long-term observation is conducted to diagnose myopia regression, then diagnostic accuracy is improved, but time cost and observation period increase
Solution Approach 1:
The patent applies preliminary action by performing fundus photography and machine learning analysis before vision correction surgery to predict the risk of myopia regression. This allows patients to be identified as high-risk beforehand, enabling preventive measures or alternative treatment plans without requiring long-term postoperative observation to diagnose regression.
Solution Approach 2:
The patent replaces the mechanical/physical observation system with an information-processing system. Instead of physically observing and measuring eye changes over time, the system uses fundus photography combined with machine learning algorithms to predict regression risk, substituting direct measurement with computational analysis.
2Manufacturing precision
If vision correction surgery is performed to correct myopia, then visual acuity is improved, but risk of myopia regression and long-term complications increases
Solution Approach 1:
The system performs risk assessment before surgery by analyzing fundus photography and patient data. This preliminary evaluation identifies patients at high risk of regression, allowing surgeons to take preventive actions such as adjusting surgical parameters, selecting different surgical methods, or providing additional postoperative monitoring protocols.
Solution Approach 2:
The machine learning model provides feedback about predicted regression risk based on preoperative fundus characteristics. This feedback loop allows clinicians to adjust treatment plans based on individual patient risk profiles, improving long-term outcomes by preventing regression in high-risk patients.
3Loss of information
If traditional observation methods are used to monitor myopia regression, then comprehensive data collection is achieved, but diagnostic efficiency and cost increase
Solution Approach 1:
The patent replaces traditional mechanical observation and measurement methods with an information-based machine learning system. The system processes fundus photography and patient data through algorithms that predict regression risk, achieving comprehensive analysis without requiring prolonged observation periods or multiple follow-up visits.
Solution Approach 2:
The machine learning model acts as an intermediary between raw fundus photography data and clinical diagnosis. Instead of directly observing and measuring eye changes over time, the model processes imaging data and patient information to generate risk predictions, serving as a mediator that translates complex data into actionable clinical insights.
Data Source
AI summary
An electronic device for predicting myopia regression, which includes a memory and a processor connected with the memory to execute instructions included in the memory. The processor collects first target data of a subject and second target data of the subject, extracts a first result value as output data for a first machine learning model by using the first target data as input data for the first machine learning model, and determines whether there is a possibility of myopia regression of the subject based on the first result value.