A Segmentation-Oriented Inter-Class Transfer Method: Application to Retinal Vessel Segmentation
This work addresses data scarcity for medical image segmentation in ophthalmology, offering an incremental improvement over existing methods.
The paper tackles the problem of data scarcity in retinal vessel segmentation by proposing a patch-based two-stage transfer method that leverages inter-class similarities, achieving accuracies of 97% on DRIVE, 96.8% on STARE, and 96.77% on HRF, outperforming current methods and human observers.
Retinal vessel segmentation, as a principal nonintrusive diagnose method for ophthalmology diseases or diabetics, suffers from data scarcity due to requiring pixel-wise labels. In this paper, we proposed a convenient patch-based two-stage transfer method. First, based on the information bottleneck theory, we insert one dimensionality-reduced layer for task-specific feature space. Next, the semi-supervised clustering is conducted to select instances, from different sources databases, possessing similarities in the feature space. Surprisingly, we empirically demonstrate that images from different classes possessing similarities contribute to better performance than some same-class instances. The proposed framework achieved an accuracy of 97%, 96.8%, and 96.77% on DRIVE, STARE, and HRF respectively, outperforming current methods and independent human observers (DRIVE (96.37%) and STARE (93.39%)).