Multi-encoder self-adaptive hard attention network with maximum intensity projections for lung nodule segmentation.
Authors
Affiliations (8)
Affiliations (8)
- Department of Pathology, Stanford University, CA 94305, USA; Department of Anesthesiology, Perioperative and Pain Medicine, Stanford University, Stanford, CA 94305, USA.
- Department of Biomedical Sciences, Seoul National University Graduate School, Seoul, 03080, Republic of Korea.
- Department of Electrical and Computer Engineering, The University of Alabama, AL 35401, USA.
- Department of Electrical Engineering, Information Technology University (ITU), Lahore, Pakistan.
- Department of Cybersecurity and Digital Forensics, Naif Arab University for Security Sciences, Riyadh 11452, Saudi Arabia.
- Department of Computational Biology, Muhammad bin Zayed University of Artificial Intelligence, Abu Dhabi, United Arab Emirates.
- School of Software, Soongsil University, 369 Sangdo-Ro, Dongjak-Gu, Seoul, 06978, Republic of Korea. Electronic address: [email protected].
- Department of Computer Science and Engineering, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea.
Abstract
Accurate lung nodule segmentation is crucial for early-stage lung cancer diagnosis, as it can substantially enhance patient survival rates. Computed tomography (CT) images are widely employed for early diagnosis in lung nodule analysis. However, the heterogeneity of lung nodules, size diversity, and the complexity of the surrounding environment pose challenges for developing robust nodule segmentation methods. In this study, we propose an efficient end-to-end framework, the Multi-Encoder Self-Adaptive Hard Attention Network (MESAHA-Net), which consists of three encoding paths, an attention block, and a decoder block that assimilates CT slice patches with both forward and backward maximum intensity projection (MIP) images. This synergy affords a profound contextual understanding of lung nodules and also results in a deluge of features. To manage the profusion of features generated, we incorporate a self-adaptive hard attention mechanism guided by region of interest (ROI) masks centered on nodular regions, which MESAHA-Net autonomously produces. The network sequentially undertakes slice-by-slice segmentation, emphasizing nodule regions to produce precise three-dimensional (3D) segmentation. The proposed framework has been comprehensively evaluated on the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset, the largest publicly available dataset for lung nodule segmentation. The results demonstrate that our approach is highly robust across various lung nodule types, outperforming previous state-of-the-art techniques in terms of segmentation performance and computational complexity, making it suitable for real-time clinical implementation of artificial intelligence (AI)-driven diagnostic tools.