Attention Res-UNet with Guided Decoder for semantic segmentation of brain tumors

• Published in: Biomedical signal processing and control Vol. 71; p. 103077
• Main Authors: Maji, Dhiraj, Sigedar, Prarthana, Singh, Munendra
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.01.2022
• Subjects: Attention gates; Brain tumor segmentation; Guided decoder; MRI; Res-UNet; Res-UNet; Brain tumor segmentation; Attention gates; MRI; Guided decoder
• ISSN: 1746-8094; 1746-8108
• DOI: 10.1016/j.bspc.2021.103077

[Display omitted] •Guided decoder supervises the learning process and produces improved features.•Weighted guided loss improves the prediction capabilities of the decoder layers.•Hybrid network architecture inculcates Attention Gates with a backbone of Res-UNet.•Evaluation is done on the High-Grade Glioma data of BraTS 2019. The automatic segmentation of brain tumors in Magnetic Resonance Imaging (MRI) plays a major role in accurate diagnosis and treatment planning. The present study proposes a new deep learning generator architecture called Attention Res-UNet with Guided Decoder (ARU-GD) for the segmentation of brain tumors. The proposed generator architecture have the capability to explicitly guide the learning process of each decoder layer. The individual loss function to each decoder layer helps to supervise the learning process of each layer in the decoder and thereby enables them to generate better feature maps. The attention gates in the generator focuses on the activation of relevant information instead of allowing all information to pass through the skip connections in the Res-UNet. Our model performed well in comparison to the baseline models i.e. UNet, Res-UNet, and Res-UNet with attention gates. The proposed ARU-GD is compared with popular deep learning models VGG-Net, MobileNet, QuickNAT, DenseNet and XceptionNet, and BraTS 2019 leaderboard models. The proposed ARU-GD has achieved Dice Scores of 0.911, 0.876 and 0.801 and mean IoU of 0.838, 0.781 and 0.668 on the whole tumor, tumor core and enhancing tumor respectively on unseen High-Grade Glioma test data. The implementation code is available on the following Github link.