2021/08/05 by Alexander MacLean, Saad Abbasi, MacLean, Alexander +15 · 1 citation
Computer Science · Medicine · #Artificial Intelligence (cs.AI) #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Machine Learning (cs.LG) #Seismology and Earthquake Studies #Ultrasound in Clinical Applications #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2108.03131
openalex publication_date 2021/08/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The Coronavirus Disease 2019 (COVID-19) pandemic has impacted many aspects of\nlife globally, and a critical factor in mitigating its effects is screening\nindividuals for infections, thereby allowing for both proper treatment for\nthose individuals as well as action to be taken to prevent further spread of\nthe virus. Point-of-care ultrasound (POCUS) imaging has been proposed as a\nscreening tool as it is a much cheaper and easier to apply imaging modality\nthan others that are traditionally used for pulmonary examinations, namely\nchest x-ray and computed tomography. Given the scarcity of expert radiologists\nfor interpreting POCUS examinations in many highly affected regions around the\nworld, low-cost deep learning-driven clinical decision support solutions can\nhave a large impact during the on-going pandemic. Motivated by this, we\nintroduce COVID-Net US, a highly efficient, self-attention deep convolutional\nneural network design tailored for COVID-19 screening from lung POCUS images.\nExperimental results show that the proposed COVID-Net US can achieve an AUC of\nover 0.98 while achieving 353X lower architectural complexity, 62X lower\ncomputational complexity, and 14.3X faster inference times on a Raspberry Pi.\nClinical validation was also conducted, where select cases were reviewed and\nreported on by a practicing clinician (20 years of clinical practice)\nspecializing in intensive care (ICU) and 15 years of expertise in POCUS\ninterpretation. To advocate affordable healthcare and artificial intelligence\nfor resource-constrained environments, we have made COVID-Net US open source\nand publicly available as part of the COVID-Net open source initiative.\n