Tunable Wavelet Unit based Convolutional Neural Network in Optical Coherence Tomography Analysis Enhancement for Classifying Type of Epiretinal Membrane Surgery

2025-07-01

AI Summary

This study addresses the challenge of automatically differentiating epiretinal membrane (ERM) peeling alone from combined ERM and internal limiting membrane (ILM) peeling using postoperative optical coherence tomography (OCT) central B-scan images. We propose an enhanced ResNet18 architecture incorporating adaptive wavelet units—OrthLatt-UwU and PR-Relax-UwU—integrated for the first time into downsampling, strided convolution, and pooling layers. Preprocessing includes energy-based cropping and wavelet denoising to strengthen surgery-type–specific feature representation. Evaluated on a clinical OCT dataset, the model achieves 78% classification accuracy, substantially surpassing the 50% average diagnostic accuracy of ophthalmologists. This work introduces the first interpretable, trainable wavelet-enhanced deep learning framework for macular surgery type identification, offering a novel foundation for clinical decision support in vitreoretinal surgery.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

In this study, we developed deep learning-based method to classify the type of surgery performed for epiretinal membrane (ERM) removal, either internal limiting membrane (ILM) removal or ERM-alone removal. Our model, based on the ResNet18 convolutional neural network (CNN) architecture, utilizes postoperative optical coherence tomography (OCT) center scans as inputs. We evaluated the model using both original scans and scans preprocessed with energy crop and wavelet denoising, achieving 72% accuracy on preprocessed inputs, outperforming the 66% accuracy achieved on original scans. To further improve accuracy, we integrated tunable wavelet units with two key adaptations: Orthogonal Lattice-based Wavelet Units (OrthLatt-UwU) and Perfect Reconstruction Relaxation-based Wavelet Units (PR-Relax-UwU). These units allowed the model to automatically adjust filter coefficients during training and were incorporated into downsampling, stride-two convolution, and pooling layers, enhancing its ability to distinguish between ERM-ILM removal and ERM-alone removal, with OrthLattUwU boosting accuracy to 76% and PR-Relax-UwU increasing performance to 78%. Performance comparisons showed that our AI model outperformed a trained human grader, who achieved only 50% accuracy in classifying the removal surgery types from postoperative OCT scans. These findings highlight the potential of CNN based models to improve clinical decision-making by providing more accurate and reliable classifications. To the best of our knowledge, this is the first work to employ tunable wavelets for classifying different types of ERM removal surgery.

Problem

• Classify epiretinal membrane surgery type using OCT scans
• Enhance accuracy with tunable wavelet units in CNN
• Improve clinical decision-making for ERM removal classification

Innovation

• Tunable wavelet units enhance CNN accuracy
• Orthogonal Lattice and Perfect Reconstruction adaptations
• Postoperative OCT scans preprocessing improves classification