Fundus image segmentation system based on multi-stage encoding alignment and lesion perception

Abstract

The application discloses a fundus image segmentation system based on multi-stage coding alignment and lesion perception, which comprises a double-branch encoder containing a multi-scale encoder and a wavelet decomposition encoder, extracts multi-scale global semantic features, separates low-frequency topological structure and high-frequency detailed texture through wavelet decomposition, and respectively models and fuses them to form multi-level coding fusion features complementary in structure and details; after adaptive enhancement of the fusion features by a lesion perception expert module, the decoder restores the spatial resolution step by step to obtain the segmentation result. In the training stage, a double-network semi-supervised mutual learning framework is adopted, a multi-stage coding alignment module is introduced, the coding features of the historical stages are averaged as the alignment target by a sliding window memory storage unit, the fusion features are aligned step by step by cosine similarity, and multi-scale feature alignment loss and cross pseudo-supervision loss are combined to optimize, so as to enhance the ability to depict the retinal layered topological structure and lesion boundary under limited labeled data.

Images