Computationally-efficient vision transformer for medical image semantic segmentation via dual pseudo-label supervision

Conference item

Ubiquitous accumulation of large volumes of data, and in- creased availability of annotated medical data in particular, has made it possible to show the many and varied benefits of deep learning to the semantic segmentation of medical im- ages. Nevertheless, data access and annotation come at a high cost in clinician time. The power of Vision Transformer (ViT) is well-documented for generic computer vision tasks involv- ing millions of images of every day objects, of which only relatively few have been annotated. Its translation to rela- tively more modest (i.e. thousands of images of) medical data is not immediately straightforward. This paper presents prac- tical avenues for training a Computationally-Efficient Semi- Supervised Vision Transformer (CESS-ViT) for medical im- age segmentation task. We propose a pure self-attention-based image segmenta- tion network which requires only limited computational re- sources. Additionally, we develop a dual pseudo-label super- vision scheme for use with semi-supervision in a ViT. Our method has been evaluated on a publicly available cardiac MRI dataset with direct comparison against other semi-supervised methods. Our results illustrate the proposed ViT-based semi-supervised method outperforms the existing methods in the semantic segmentation of cardiac ventricles.

Authors: Wang, Z; Dong, N; Voiculescu, ID

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: IEEE
• Host title: Proceedings of the 29th IEEE International Conference on Image Processing (ICIP 2022)
• Pages: 1961-1965
• Publication date: 2022-10-18
• Acceptance date: 2022-06-20
• Event title: 29th IEEE International Conference on Image Processing (ICIP 2022)
• Event location: Bordeaux, France
• Event website: https://2022.ieeeicip.org/
• Event start date: 2022-10-16
• Event end date: 2022-10-19
• DOI: 10.1109/ICIP46576.2022.9897482
• EISSN: 2381-8549
• ISSN: 1522-4880
• Language: English
• Keywords: FFR