Biomedical image segmentation at point of care

René Groh, Stephan Dürr, Anne Schützenberger, Marion Semmler, Andreas M Kist

Keywords: Deep Learning, Endoscopy, Semantic Segmentation, Point of Care

TL;DR: Deep neural networks optimized for clinical use show consistent performance in an endoscopic segmentation task across a two-year period.

Abstract: Deep Learning has a large impact on medical image analysis and lately has been adopted for clinical use at the point of care. However, there is only a small number of reports of long-term studies that show the performance of deep neural networks (DNNs) in such a clinical environment. We have collected the video footage for two years from an AI-powered laryngeal high-speed videoendoscopy imaging system. In this study, we aimed to measure the performance of the deployed DNN, which is designed for a specific task, laryngeal glottis segmentation. We first evaluated the image quality of the video footage as the image quality has a major impact on segmentation performance and found that the image quality is stable across time. Next, we determined the DNN segmentation performance on lossy and lossless compressed data and measured if manual correction of segmentation masks is needed. We found that lossy and lossless compression are on par for glottis segmentation, however, lossless compression provides significantly superior image quality. Lastly, we employed continual learning strategies to continuously incorporate new data to the DNN to remove segmentation artefacts that occur in 9% of recordings. With modest manual intervention, we were able to largely alleviate these segmentation artefacts by up to 81%. We believe that our suggested deep learning-enhanced laryngeal imaging platform consistently provides clinically sound results, and together with our proposed continual learning scheme will have a long-lasting impact in the future of laryngeal imaging.

Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: validation/application paper

Primary Subject Area: Application: Endoscopy

Secondary Subject Area: Segmentation

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.