IOL Power Prediction Using Intraoperative Refraction Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional IOL calculators rely on pre-operative data and are prone to errors, leading to significant inaccuracies in post-operative refractive outcomes, affecting a substantial number of patients and increasing surgical costs and time.
Innovation Solution
A machine learning-based system that incorporates pre-operative and intra-operative measurements, using a determination engine to generate a predictive feature set and refine models for accurate post-operative lens settlement prediction, reducing errors to near zero.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional calculators are used for IOL power prediction, then the surgical process is simpler and faster, but the prediction accuracy deteriorates with mean error of ±0.343 D
Solution Approach 1:
The system segments the prediction process into multiple independent modules: pre-operative measurement module, intra-operative measurement module, and prediction module. Each module handles specific measurement tasks (e.g., axial length, corneal curvature, anterior chamber depth) and processes data independently before integration, allowing complex predictions to be broken down into manageable components that can be developed, tested, and maintained separately while achieving higher overall accuracy
Solution Approach 2:
The system introduces an intermediary determination engine that acts as a mediator between raw measurement data and final IOL power predictions. This engine processes pre-operative and intra-operative measurements through multiple algorithms, reconciles data from different sources, and produces refined predictions. The intermediary layer isolates the complexity of data integration and algorithm coordination from both the measurement devices and the final decision-making process
2Measurement precision
If conventional calculators with interoperative data are used, then lens position estimation is provided, but the surgical time and costs increase
Solution Approach 1:
The system performs preliminary measurements and calculations during the pre-operative phase, capturing baseline ocular parameters (axial length, keratometry, anterior chamber depth) before surgery. This advance data collection allows the intra-operative prediction engine to work with pre-processed information, reducing the computational burden during surgery and enabling faster predictions without sacrificing accuracy. The preliminary action of data collection and initial processing eliminates the need for time-consuming measurements during the surgical procedure
3Ease of operation
If conventional calculators are used, then the process is simpler to operate, but the prediction error affects 5% of patients with more than 1.0 D deviation
Solution Approach 1:
The system implements self-service through automated data collection, processing, and prediction generation. The determination engine automatically ingests pre-operative and intra-operative measurements, selects appropriate algorithms, performs calculations, and outputs IOL power recommendations without requiring manual intervention. This automation maintains ease of operation while significantly improving reliability by eliminating human calculation errors and ensuring consistent application of predictive algorithms across all cases
Solution Approach 2:
The system incorporates feedback mechanisms where actual post-operative outcomes are fed back into the determination engine to continuously refine prediction algorithms. By comparing predicted versus actual refractive outcomes, the system adjusts algorithm parameters and weighting factors to improve accuracy over time. This feedback loop enhances reliability while maintaining operational simplicity, as the refinement process occurs automatically in the background without adding complexity to the user interface
Data Source
AI summary
Described are implementations of systems and methods for an improved machine learning-based system that incorporates pre-operative and intraoperative measurements captured during surgery, as well as additional patient-specific data, to provide an individualized, highly accurate post-operative manifest refraction prediction. According to some embodiments, a determination engine generates a predictive feature set of one or more predictors associated with diagnostic measurements of one or more eyes and performs a recursive selection operation using one or more combinations within the predictive feature set and one or more models to produce a most predictive subset, the most predictive subset having a highest prediction accuracy among other predictive subsets for post-operative manifest refraction. The determination engine generates a determination model by refining and retraining the one or more models of the recursive selection operation utilizing the most predictive subset.