Computer-aided detection (CAD) approaches have shown promising results for early esophageal cancer detection using Volumetric Laser Endoscopy (VLE) imagery. However, the relatively slow and computationally costly tissue segmentation employed in these approaches hamper their clinical applicability. In this paper, we propose to reframe the 2D tissue segmentation problem into a 1D tissue boundary detection problem. Instead of using an encoder-decoder architecture, we propose to follow the tissue boundary using a Recurrent Neural Network (RNN), exploiting the spatio-temporal relations within VLE frames. We demonstrate a near state-of-the-art performance using 18 times less floating point operations, enabling real-time execution in clinical practice.
|Title of host publication||Medical Imaging 2021|
|Subtitle of host publication||Computer-Aided Diagnosis|
|Editors||Maciej A. Mazurowski, Karen Drukker|
|Publication status||Published - 2021|
|Event||Medical Imaging 2021: Computer-Aided Diagnosis - Virtual, Online, United States|
Duration: 15 Feb 2021 → 19 Feb 2021
|Name||Proceedings of SPIE|
|Conference||Medical Imaging 2021: Computer-Aided Diagnosis|
|Period||15/02/21 → 19/02/21|
Bibliographical notePublisher Copyright:
© COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.
Copyright 2021 Elsevier B.V., All rights reserved.
- Barrett's esophagus
- Computer aided detection
- Deep learning
- Recurrent neural network
- Volumetric laser endomicroscopy