Embedded Files
Brain tumour segmentation is critical in medical image analysis, facilitating diagnosis and treatment planning in neurosurgery. Brain tumour segmentation models using supervised learning show robust results in medical imaging, however, they have the drawback of requiring a substantial amount of annotated data to perform effectively. In addition, accurately detecting the boundaries of tumour subregions is crucial in fine-grained segmentation. However, existing studies lack effective methods for achieving this. We propose a novel approach using a unique dual-decoder architecture, focusing on edge identification and enhancing segmentation accuracy. Utilizing a dual-decoder 3D-UNet model, we prioritize accuracy and fine-grained details in tumour segmentation and introduce an additional tumour edge detection task as well to the model, aiming to move beyond traditional single-decoder approaches. Incorporating a 3D SimSiam network as the self-supervised pretraining technique, we aim to address the limitation of annotated data and enhance the segmentation accuracy. Our model surpasses many supervised variants of U-net architectures and self-supervised approaches, highlighting the importance of edge detection in tumour segmentation. The proposed approach enhances segmentation accuracy by showing an accuracy of 98.1% and provides critical boundary details for clinical decision-making. Visualizations of segmentation and edge masks further validate the effectiveness of the proposed method.
Team: Mr. Theshan Wijerathne, Mr. Deshan Wickramasinghe, Mr. Dasith Samarasinghe
Page updated
Report abuse