@inproceedings{c7bcc0f5423448c7987b2904af1b3304,
title = "UNet-eVAE: Iterative Refinement Using VAE Embodied Learning for Endoscopic Image Segmentation",
abstract = "While endoscopy is routinely used for surveillance, high operator dependence demands robust automated image analysis methods. Automated segmentation of region-of-interest (ROI) that includes lesions, inflammations, and instruments can serve to cope with the operator dependence problem in this field. Most supervised methods are developed by fitting models on the available ground truth mask samples only. This work proposes a joint training approach using the UNet coupled with a variational auto-encoder (VAE) to improve endoscopic image segmentation by exploiting original samples, predicted masks and ground truth masks. In the proposed UNet-eVAE, VAE utilises the masks to constrain ROI-specific feature representations for reconstruction as an auxiliary task. The fine-grained spatial information from VAE is fused with the UNet decoder to enrich the feature representations and improve segmentation performance. Our experimental results on both colonoscopy and ureteroscopy datasets demonstrate that the proposed architecture can learn robust representations and generalise segmentation performance on unseen samples while improving the baseline.",
author = "Soumya Gupta and Sharib Ali and Xu Zhang and Binod Bhattarai and Jens Rittscher",
note = "We would like to thank Boston Scientific for funding this project (Grant No: DFR04690). SG and BT are funded by BSC, BB is funded by EndoMapper Horizon 2020 FET (GA 863146), SA and JR were supported by the NIHR Oxford Biomedical Research Centre.",
year = "2022",
month = dec,
day = "16",
doi = "10.1007/978-3-031-21014-3_17",
language = "English",
isbn = "978-3-031-21013-6",
series = "Lecture Notes in Computer Science",
publisher = "Springer Nature",
pages = "161--170",
booktitle = "Machine Learning in Medical Imaging",
address = "United States",
}