Intraocular Lens Position Prediction Using 3D Interpolation Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting the postoperative horizontal depth position of intraocular lenses in the eye are inaccurate due to oversimplifications and assumptions, leading to suboptimal correction of visual defects and potential surgical complications.
Innovation Solution
A method and device that utilize a three-dimensional coordinate system with selected main reference points and edge connecting lines to generate an uneven prediction network, allowing for a more precise prediction of the postoperative horizontal depth position based on preoperative parameters such as anterior chamber depth and eye length.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional simplified models and assumptions are used for predicting intraocular lens position, then the calculation process is simple and quick, but the prediction accuracy is insufficient
Solution Approach 1:
The prediction model is segmented into multiple components: a three-dimensional coordinate system with x, y, z axes representing different anatomical parameters, main reference points at specific coordinate locations, and edge connecting lines forming a prediction network. This segmentation allows complex predictions to be broken down into manageable geometric relationships between discrete points and lines.
Solution Approach 2:
The invention transitions from traditional two-dimensional or simplified one-dimensional prediction models to a comprehensive three-dimensional coordinate system. The main reference points are positioned at specific (x, y, z) coordinates, and edge connecting lines are defined by coordinate pairs, enabling spatial prediction of intraocular lens position in three dimensions rather than along a single axis.
2Reliability
If accurate prediction of intraocular lens position is achieved through detailed modeling, then the corrective effect is improved, but the computational complexity and time increase
Solution Approach 1:
The prediction network is pre-established using main reference points and edge connecting lines before actual intraocular lens positioning is required. The three-dimensional coordinate system and prediction network structure are prepared in advance, allowing rapid interpolation and prediction calculations to be performed once when clinical decisions are needed, rather than requiring complex computations during the surgical planning phase.
3Adaptability or versatility
If a uniform prediction model is used for all eyes, then the model is simple to implement, but it cannot account for individual variations in eye anatomy
Solution Approach 1:
The prediction model incorporates local anatomical variations by positioning main reference points at specific coordinate locations that correspond to individual eye characteristics. The edge connecting lines are defined by local coordinate pairs that adapt to the specific geometry of each patient's eye anatomy, allowing the model to provide localized accurate predictions rather than applying a uniform approach to all cases.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The invention relates to a method and a device for preoperatively predicting a postoperative horizontal depth (postope IOLE, postope IOLE*) of an intraocular lens (14, 14') in a patient's eye (1). Parameter values are provided, said parameters including a first parameter as the depth (preopACD) of an anterior chamber (7) of human eyes, a second parameter as a horizontal length (preopAL) of human eyes, and a third parameter as a postoperative position of intraocular lenses in human eyes. Specific triplets of values are selected from said values as main interpolation nodes (18 to 33), each triplet of values comprising a value for each of the three aforementioned parameters, and the main interpolation nodes (18 to 33) are entered into a three-dimensional coordinate system that is spanned by the three parameters. The main interpolation nodes (18 to 33) are connected by edge connecting lines (34 to 57), and a predictive network (58) that is unevenly formed in the three-dimensional coordinate system is generated at least using the positions of the edge connecting lines (34 to 57) in order to predict the postoperative horizontal depth (postopeIOLE, postopeIOLE*) of the intraocular lens (14, 14') in the patient's eye (1).