IOL Power Prediction Using Lens Settlement and Surgical Measurements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional calculators for intraocular lens (IOL) implantation in cataract surgery are inaccurate, leading to significant refractive errors and dissatisfaction among patients, with a mean prediction error of ±0.343 diopters and a standard deviation of ±0.343 diopters, affecting a large number of patients post-surgery.
Innovation Solution
A machine learning-based system that utilizes pre-operative and intra-operative measurements, including high-definition measurements from laser-based surgical systems, to predict post-operative lens settlement and improve refractive outcomes by generating a predictive feature set and determining a most predictive subset using a determination engine, which accounts for post-operative lens settlement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional calculators are used for IOL power prediction, then the surgery cost and time are reduced, but the prediction accuracy deteriorates with a mean error of ±0.343 diopters
Solution Approach 1:
The system segments the prediction process into multiple independent components: pre-operative biometry measurements, intraoperative OCT measurements, and machine learning model processing. This segmentation allows each component to be optimized independently while maintaining overall system manageability and accuracy.
Solution Approach 2:
The patent introduces a machine learning determination model as an intermediary between raw biometry measurements and final IOL power calculation. This intermediary processes multiple measurement parameters simultaneously and accounts for lens settlement effects, achieving ±0.15 diopter accuracy without requiring complex manual calculations.
2Reliability
If conventional calculators with interoperative data and lens estimations are used, then lens position recommendations are provided, but the surgery cost and time increase
Solution Approach 1:
The system performs preliminary biometry measurements pre-operatively using optical biometry devices, capturing axial length, corneal curvature, and anterior chamber depth before surgery. This preliminary data collection enables the machine learning model to make accurate predictions without requiring extensive intraoperative measurements, reducing surgery time while maintaining reliability.
Solution Approach 2:
The patent implements continuous measurement and prediction throughout the surgical workflow. The machine learning determination model continuously processes available biometry data during surgery, providing real-time IOL power recommendations without interrupting the surgical flow or requiring separate calculation steps.
3Productivity
If conventional calculators are used, then the calculation process is simplified, but the refractive outcome accuracy deteriorates affecting 20 million cataract surgeries annually
Solution Approach 1:
The machine learning determination model automatically processes biometry measurements and generates IOL power predictions without requiring manual calculation steps. The system self-adjusts for lens settlement effects and provides final recommendations, enabling high-volume surgery capacity while maintaining ±0.15 diopter precision across 20 million annual cataract procedures.
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.