Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

OUCopula: Bi-Channel Multi-Label Copula-Enhanced Adapter-Based CNN for Myopia Screening Based on OU-UWF Images

Yang Li, Qiuyi Huang, Chong Zhong, Danjuan Yang, Meiyan Li, A. H. Welsh, Aiyi Liu, Bo Fu, Catherien C. Liu, Xingtao Zhou

TL;DR

A novel bi-channel multi-label CNN which can take bi-channel image inputs subject to both high correlation and heterogeneity, and incorporate correlation information between continuous output labels (using a copula) and the potential extension of the bi-channel model to a multi-channel paradigm and the generalizability of OUCopula across various backbone CNNs.

Abstract

Myopia screening using cutting-edge ultra-widefield (UWF) fundus imaging is potentially significant for ophthalmic outcomes. Current multidisciplinary research between ophthalmology and deep learning (DL) concentrates primarily on disease classification and diagnosis using single-eye images, largely ignoring joint modeling and prediction for Oculus Uterque (OU, both eyes). Inspired by the complex relationships between OU and the high correlation between the (continuous) outcome labels (Spherical Equivalent and Axial Length), we propose a framework of copula-enhanced adapter convolutional neural network (CNN) learning with OU UWF fundus images (OUCopula) for joint prediction of multiple clinical scores. We design a novel bi-channel multi-label CNN that can (1) take bi-channel image inputs subject to both high correlation and heterogeneity (by sharing the same backbone network and employing adapters to parameterize the channel-wise discrepancy), and (2) incorporate correlation information between continuous output labels (using a copula). Solid experiments show that OUCopula achieves satisfactory performance in myopia score prediction compared to backbone models. Moreover, OUCopula can far exceed the performance of models constructed for single-eye inputs. Importantly, our study also hints at the potential extension of the bi-channel model to a multi-channel paradigm and the generalizability of OUCopula across various backbone CNNs.

OUCopula: Bi-Channel Multi-Label Copula-Enhanced Adapter-Based CNN for Myopia Screening Based on OU-UWF Images

TL;DR

A novel bi-channel multi-label CNN which can take bi-channel image inputs subject to both high correlation and heterogeneity, and incorporate correlation information between continuous output labels (using a copula) and the potential extension of the bi-channel model to a multi-channel paradigm and the generalizability of OUCopula across various backbone CNNs.

Abstract

Myopia screening using cutting-edge ultra-widefield (UWF) fundus imaging is potentially significant for ophthalmic outcomes. Current multidisciplinary research between ophthalmology and deep learning (DL) concentrates primarily on disease classification and diagnosis using single-eye images, largely ignoring joint modeling and prediction for Oculus Uterque (OU, both eyes). Inspired by the complex relationships between OU and the high correlation between the (continuous) outcome labels (Spherical Equivalent and Axial Length), we propose a framework of copula-enhanced adapter convolutional neural network (CNN) learning with OU UWF fundus images (OUCopula) for joint prediction of multiple clinical scores. We design a novel bi-channel multi-label CNN that can (1) take bi-channel image inputs subject to both high correlation and heterogeneity (by sharing the same backbone network and employing adapters to parameterize the channel-wise discrepancy), and (2) incorporate correlation information between continuous output labels (using a copula). Solid experiments show that OUCopula achieves satisfactory performance in myopia score prediction compared to backbone models. Moreover, OUCopula can far exceed the performance of models constructed for single-eye inputs. Importantly, our study also hints at the potential extension of the bi-channel model to a multi-channel paradigm and the generalizability of OUCopula across various backbone CNNs.
Paper Structure (16 sections, 6 equations, 5 figures, 1 algorithm)

This paper contains 16 sections, 6 equations, 5 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methods
  4. Residual adapter modules
  5. Copula modeling
  6. Multi-label regression
  7. Gaussian copula
  8. Copula loss
  9. Choice of backbone models
  10. End-to-end OUCopula
  11. Experiments
  12. Dataset
  13. Experiment Setup
  14. Results
  15. Discussion
  16. ...and 1 more sections

Figures (5)

  • Figure 1: UWF fundus images of one patient indicating interocular asymmetries in myopia status
  • Figure 2: Heatmap of correlation between SE and AL from both eyes
  • Figure 3: (a) Architecture of proposed OUCopula; (b)Detailed architecture of ResNet backbone with adapter modules.
  • Figure 4: Box plots of 5-fold CV prediction performances of OUCopula and baselines for all sub-results.
  • Figure 5: The prediction performances (in average among 5 folds) of OUCopula and baselines for all sub-results.