Image Magnification Network for Vessel Segmentation in OCTA Images
This work addresses the segmentation of microvasculature in OCTA images, which is crucial for medical diagnosis, but it appears incremental as it modifies the U-Net architecture for better detail capture.
The paper tackles the problem of retinal vessel segmentation in OCTA images, particularly the challenge of thin and dense capillary structures, by proposing an image magnification network (IMN) that achieves an average dice score of 90.2% on three datasets, outperforming existing methods.
Optical coherence tomography angiography (OCTA) is a novel non-invasive imaging modality that allows micron-level resolution to visualize the retinal microvasculature. The retinal vessel segmentation in OCTA images is still an open problem, and especially the thin and dense structure of the capillary plexus is an important challenge of this problem. In this work, we propose a novel image magnification network (IMN) for vessel segmentation in OCTA images. Contrary to the U-Net structure with a down-sampling encoder and up-sampling decoder, the proposed IMN adopts the design of up-sampling encoding and then down-sampling decoding. This design is to capture more low-level image details to reduce the omission of small structures. The experimental results on three open OCTA datasets show that the proposed IMN with an average dice score of 90.2% achieves the best performance in vessel segmentation of OCTA images. Besides, we also demonstrate the superior performance of IMN in cross-field image vessel segmentation and vessel skeleton extraction.