Multimodal Myopia Prediction Using Ocular Imaging and Clinical Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current eye health assessment systems struggle to efficiently detect and predict the progression of myopia, particularly in pediatric settings, due to the shortage of ophthalmic experts and the need for timely treatment plans.
Innovation Solution
A system utilizing a processor with non-transitory memory that executes instructions for receiving and processing ocular region images, clinical data, and external factors, employing adaptive filtering and convolutional neural networks for ocular characteristic extraction, and a multimodal fusion module to predict myopia onset and progression, with dynamic embedding adjustments and feedback mechanisms for enhanced accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional manual image analysis by human experts is used, then diagnostic accuracy is maintained, but system productivity and efficiency deteriorate due to shortage of experts and increasing demand
Solution Approach 1:
The patent replaces manual human expert analysis with an automated deep learning-based image analysis system. The neural network model processes retinal images to detect myopia and other ocular conditions, eliminating the bottleneck of human expert shortage while maintaining diagnostic capability through sophisticated algorithmic analysis
Solution Approach 2:
The system creates a virtual copy of human expert diagnostic capability through trained neural network models. These models learn from extensive training data and replicate the diagnostic reasoning processes of experts, enabling scalable deployment without requiring physical duplication of human expertise
2Loss of information
If comprehensive eye examinations are conducted to detect current condition and predict future progression, then diagnostic completeness is improved, but examination time and patient burden increase
Solution Approach 1:
The system performs preliminary analysis by extracting and analyzing multiple ocular characteristics (cup-disc ratio, optic disc size, retinal nerve fiber layer thickness, axial length) during the initial examination. This preliminary data extraction and prediction of myopia progression occurs automatically as part of the standard workflow, eliminating the need for separate dedicated prediction examinations
Solution Approach 2:
The imaging system serves multiple functions simultaneously: it captures retinal images, extracts quantitative ocular characteristics, detects current myopia status, and predicts future progression. This multi-functional approach consolidates what would traditionally require separate examination procedures into a single comprehensive assessment
3Measurement precision
If multiple ocular characteristics and external factors are integrated for holistic analysis, then prediction accuracy is improved, but system complexity increases
Solution Approach 1:
The patent segments the complex analysis task into distinct functional modules: image acquisition, characteristic extraction (cup-disc ratio, optic disc size, RNFL thickness, axial length), myopia detection, and progression prediction. Each module processes specific data independently before integrating results, making the overall complex system manageable and maintainable
Solution Approach 2:
The system merges multiple data sources including ocular imaging characteristics, patient demographic data, and external factors (outdoor activity, screen time, parental myopia history) into a unified prediction model. This integration combines diverse data types through standardized processing pipelines that handle each factor systematically before consolidating results
Data Source
AI summary
According to an embodiment, disclosed is a system comprising a processor wherein the processor is configured to receive an input data comprising an image of an ocular region of a user, clinical data of the user, and external factors; extract, using an image processing module comprising adaptive filtering techniques, ocular characteristics, combine, using a multimodal fusion module, the input data to determine a holistic health embedding; detect, based on a machine learning model and the holistic health embedding, a first output comprising likelihood of myopia, and severity of myopia; predict, based on the machine learning model and the holistic health embedding, a second output comprising an onset of myopia and a progression of myopia in the user; and wherein the machine learning model is a pre-trained model; and wherein the system is configured for myopia prognosis powered by multimodal data.