Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multiscale Color Guided Attention Ensemble Classifier for Age-Related Macular Degeneration using Concurrent Fundus and Optical Coherence Tomography Images

Pragya Gupta, Subhamoy Mandal, Debashree Guha, Debjani Chakraborty

TL;DR

The paper addresses AMD diagnosis by leveraging complementary information from Fundus and OCT images through a multimodal, multiscale color-space framework. The proposed MCGAEc model uses modality-specific encoders that transform Fundus data into two color spaces (YCbCr and HSV) and process OCT data in grayscale across multiple scales, all feeding pre-trained VGG16 backbones. Self-attention refines path-specific features before concatenation and Random Forest classification, enabling robust AMD staging. Experimental results on the Project Macula dataset show state-of-the-art performance (AUC up to 0.994) and clear gains over single-modality and other SOTA methods, underscoring the value of cross-modality, multiscale, and attention-guided feature fusion for retinal disease diagnosis and potential clinical deployment.

Abstract

Automatic diagnosis techniques have evolved to identify age-related macular degeneration (AMD) by employing single modality Fundus images or optical coherence tomography (OCT). To classify ocular diseases, fundus and OCT images are the most crucial imaging modalities used in the clinical setting. Most deep learning-based techniques are established on a single imaging modality, which contemplates the ocular disorders to a specific extent and disregards other modality that comprises exhaustive information among distinct imaging modalities. This paper proposes a modality-specific multiscale color space embedding integrated with the attention mechanism based on transfer learning for classification (MCGAEc), which can efficiently extract the distinct modality information at various scales using the distinct color spaces. In this work, we first introduce the modality-specific multiscale color space encoder model, which includes diverse feature representations by integrating distinct characteristic color spaces on a multiscale into a unified framework. The extracted features from the prior encoder module are incorporated with the attention mechanism to extract the global features representation, which is integrated with the prior extracted features and transferred to the random forest classifier for the classification of AMD. To analyze the performance of the proposed MCGAEc method, a publicly available multi-modality dataset from Project Macula for AMD is utilized and compared with the existing models.

Multiscale Color Guided Attention Ensemble Classifier for Age-Related Macular Degeneration using Concurrent Fundus and Optical Coherence Tomography Images

TL;DR

The paper addresses AMD diagnosis by leveraging complementary information from Fundus and OCT images through a multimodal, multiscale color-space framework. The proposed MCGAEc model uses modality-specific encoders that transform Fundus data into two color spaces (YCbCr and HSV) and process OCT data in grayscale across multiple scales, all feeding pre-trained VGG16 backbones. Self-attention refines path-specific features before concatenation and Random Forest classification, enabling robust AMD staging. Experimental results on the Project Macula dataset show state-of-the-art performance (AUC up to 0.994) and clear gains over single-modality and other SOTA methods, underscoring the value of cross-modality, multiscale, and attention-guided feature fusion for retinal disease diagnosis and potential clinical deployment.

Abstract

Automatic diagnosis techniques have evolved to identify age-related macular degeneration (AMD) by employing single modality Fundus images or optical coherence tomography (OCT). To classify ocular diseases, fundus and OCT images are the most crucial imaging modalities used in the clinical setting. Most deep learning-based techniques are established on a single imaging modality, which contemplates the ocular disorders to a specific extent and disregards other modality that comprises exhaustive information among distinct imaging modalities. This paper proposes a modality-specific multiscale color space embedding integrated with the attention mechanism based on transfer learning for classification (MCGAEc), which can efficiently extract the distinct modality information at various scales using the distinct color spaces. In this work, we first introduce the modality-specific multiscale color space encoder model, which includes diverse feature representations by integrating distinct characteristic color spaces on a multiscale into a unified framework. The extracted features from the prior encoder module are incorporated with the attention mechanism to extract the global features representation, which is integrated with the prior extracted features and transferred to the random forest classifier for the classification of AMD. To analyze the performance of the proposed MCGAEc method, a publicly available multi-modality dataset from Project Macula for AMD is utilized and compared with the existing models.
Paper Structure (8 sections, 4 equations, 4 figures, 3 tables)

This paper contains 8 sections, 4 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Proposed Model
  3. Experimental Framework
  4. Dataset and Evaluation Measures
  5. Effectiveness of the Proposed MCGAEc Model
  6. Quantitative Comparison with Existing Approaches
  7. Discussion
  8. Conclusion

Figures (4)

  • Figure 1: (a) The Original Fundus image and (b) the corresponding OCT image.
  • Figure 2: The proposed framework for the combination of the Fundus and OCT images for the classification of AMD
  • Figure 3: Representation of the Fundus image in (a) RGB color space (original), (b) YCbCr color space, and (c) HSV color space.
  • Figure 4: ROC (Receiver operating characteristic) curves for the proposed method, (a) multiscale color space with single modality comparison, (b) fusion of the multiscale color spaces, (c) combination of multi-modality fundus and OCT images for the multi-class classification of AMD.