Deep Learning for Detection and Localization of B-Lines in Lung Ultrasound
Authors:
Ruben T. Lucassen,
Mohammad H. Jafari,
Nicole M. Duggan,
Nick Jowkar,
Alireza Mehrtash,
Chanel Fischetti,
Denie Bernier,
Kira Prentice,
Erik P. Duhaime,
Mike **,
Purang Abolmaesumi,
Friso G. Heslinga,
Mitko Veta,
Maria A. Duran-Mendicuti,
Sarah Frisken,
Paul B. Shyn,
Alexandra J. Golby,
Edward Boyer,
William M. Wells,
Andrew J. Goldsmith,
Tina Kapur
Abstract:
Lung ultrasound (LUS) is an important imaging modality used by emergency physicians to assess pulmonary congestion at the patient bedside. B-line artifacts in LUS videos are key findings associated with pulmonary congestion. Not only can the interpretation of LUS be challenging for novice operators, but visual quantification of B-lines remains subject to observer variability. In this work, we inve…
▽ More
Lung ultrasound (LUS) is an important imaging modality used by emergency physicians to assess pulmonary congestion at the patient bedside. B-line artifacts in LUS videos are key findings associated with pulmonary congestion. Not only can the interpretation of LUS be challenging for novice operators, but visual quantification of B-lines remains subject to observer variability. In this work, we investigate the strengths and weaknesses of multiple deep learning approaches for automated B-line detection and localization in LUS videos. We curate and publish, BEDLUS, a new ultrasound dataset comprising 1,419 videos from 113 patients with a total of 15,755 expert-annotated B-lines. Based on this dataset, we present a benchmark of established deep learning methods applied to the task of B-line detection. To pave the way for interpretable quantification of B-lines, we propose a novel "single-point" approach to B-line localization using only the point of origin. Our results show that (a) the area under the receiver operating characteristic curve ranges from 0.864 to 0.955 for the benchmarked detection methods, (b) within this range, the best performance is achieved by models that leverage multiple successive frames as input, and (c) the proposed single-point approach for B-line localization reaches an F1-score of 0.65, performing on par with the inter-observer agreement. The dataset and developed methods can facilitate further biomedical research on automated interpretation of lung ultrasound with the potential to expand the clinical utility.
△ Less
Submitted 15 February, 2023;
originally announced February 2023.
