SAM-OCTA: A Fine-Tuning Strategy for Applying Foundation Model to OCTA Image Segmentation Tasks

TL;DR

SAM-OCTA introduces a novel fine-tuning approach using low-rank adaptation and prompt strategies to improve OCTA image segmentation, achieving state-of-the-art results on vessel and artery-vein segmentation tasks.

Contribution

The paper proposes SAM-OCTA, a new fine-tuning method for foundation models that enhances OCTA image segmentation, especially for local vessel and artery-vein segmentation.

Findings

Achieves state-of-the-art segmentation performance on OCTA-500 dataset.

Effectively segments local vessels and artery-vein structures.

Addresses overfitting issues in limited supervised data scenarios.

Abstract

In the analysis of optical coherence tomography angiography (OCTA) images, the operation of segmenting specific targets is necessary. Existing methods typically train on supervised datasets with limited samples (approximately a few hundred), which can lead to overfitting. To address this, the low-rank adaptation technique is adopted for foundation model fine-tuning and proposed corresponding prompt point generation strategies to process various segmentation tasks on OCTA datasets. This method is named SAM-OCTA and has been experimented on the publicly available OCTA-500 dataset. While achieving state-of-the-art performance metrics, this method accomplishes local vessel segmentation as well as effective artery-vein segmentation, which was not well-solved in previous works. The code is available at: https://github.com/ShellRedia/SAM-OCTA.