TY - JOUR
T1 - Deep Learning for Classification and Localization of COVID-19 Markers in Point-of-Care Lung Ultrasound
AU - Roy, Subhankar
AU - Menapace, Willi
AU - Oei, Sebastiaan
AU - Luijten, Ben
AU - Fini, Enrico
AU - Saltori, Cristiano
AU - Huijben, Iris
AU - Chennakeshava, Nishith
AU - Mento, Federico
AU - Sentelli, Alessandro
AU - Peschiera, Emanuele
AU - Trevisan, Riccardo
AU - Maschietto, Giovanni
AU - Torri, Elena
AU - Inchingolo, Riccardo
AU - Smargiassi, Andrea
AU - Soldati, Gino
AU - Rota, Paolo
AU - Passerini, Andrea
AU - van Sloun, Ruud J. G.
AU - Ricci, Elisa
AU - Demi, Libertario
N1 - Fondazione VRT
Caritro Deep Learning Lab
PY - 2020/8
Y1 - 2020/8
N2 - Deep learning (DL) has proved successful in medical imaging and, in the wake of the recent COVID-19 pandemic, some works have started to investigate DL-based solutions for the assisted diagnosis of lung diseases. While existing works focus on CT scans, this paper studies the application of DL techniques for the analysis of lung ultrasonography (LUS) images. Specifically, we present a novel fully-annotated dataset of LUS images collected from several Italian hospitals, with labels indicating the degree of disease severity at a frame-level, video-level, and pixel-level (segmentation masks). Leveraging these data, we introduce several deep models that address relevant tasks for the automatic analysis of LUS images. In particular, we present a novel deep network, derived from Spatial Transformer Networks, which simultaneously predicts the disease severity score associated to a input frame and provides localization of pathological artefacts in a weakly-supervised way. Furthermore, we introduce a new method based on uninorms for effective frame score aggregation at a video-level. Finally, we benchmark state of the art deep models for estimating pixel-level segmentations of COVID-19 imaging biomarkers. Experiments on the proposed dataset demonstrate satisfactory results on all the considered tasks, paving the way to future research on DL for the assisted diagnosis of COVID-19 from LUS data.
AB - Deep learning (DL) has proved successful in medical imaging and, in the wake of the recent COVID-19 pandemic, some works have started to investigate DL-based solutions for the assisted diagnosis of lung diseases. While existing works focus on CT scans, this paper studies the application of DL techniques for the analysis of lung ultrasonography (LUS) images. Specifically, we present a novel fully-annotated dataset of LUS images collected from several Italian hospitals, with labels indicating the degree of disease severity at a frame-level, video-level, and pixel-level (segmentation masks). Leveraging these data, we introduce several deep models that address relevant tasks for the automatic analysis of LUS images. In particular, we present a novel deep network, derived from Spatial Transformer Networks, which simultaneously predicts the disease severity score associated to a input frame and provides localization of pathological artefacts in a weakly-supervised way. Furthermore, we introduce a new method based on uninorms for effective frame score aggregation at a video-level. Finally, we benchmark state of the art deep models for estimating pixel-level segmentations of COVID-19 imaging biomarkers. Experiments on the proposed dataset demonstrate satisfactory results on all the considered tasks, paving the way to future research on DL for the assisted diagnosis of COVID-19 from LUS data.
KW - Image segmentation
KW - Lung
KW - Ultrasonic imaging
KW - Task analysis
KW - Pathology
KW - Imaging
KW - Diseases
KW - COVID-19
KW - lung ultrasound
KW - deep learning
KW - Coronavirus Infections/diagnostic imaging
KW - Pandemics
KW - Lung/diagnostic imaging
KW - Humans
KW - Ultrasonography/methods
KW - Deep Learning
KW - Point-of-Care Systems
KW - Betacoronavirus
KW - Image Interpretation, Computer-Assisted/methods
KW - Pneumonia, Viral/diagnostic imaging
UR - http://www.scopus.com/inward/record.url?scp=85087653114&partnerID=8YFLogxK
U2 - 10.1109/TMI.2020.2994459
DO - 10.1109/TMI.2020.2994459
M3 - Article
C2 - 32406829
SN - 0278-0062
VL - 39
SP - 2676
EP - 2687
JO - IEEE Transactions on Medical Imaging
JF - IEEE Transactions on Medical Imaging
IS - 8
M1 - 9093068
ER -