2019/06/21 by Tanveer Syeda-Mahmood, Hassan Ahmad, Syeda-Mahmood, Tanveer +21
Medicine · #COVID-19 diagnosis using AI #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Radiology practices and education #Radiomics and Machine Learning in Medical Imaging
paper · pdf · doi:10.48550/arxiv.1906.09336
openalex publication_date 2019/06/21 · openalex created_date 2022/07/22 · openalex updated_date 2026/07/28
Chest X-rays are the most common diagnostic exams in emergency rooms and\nhospitals. There has been a surge of work on automatic interpretation of chest\nX-rays using deep learning approaches after the availability of large open\nsource chest X-ray dataset from NIH. However, the labels are not sufficiently\nrich and descriptive for training classification tools. Further, it does not\nadequately address the findings seen in Chest X-rays taken in\nanterior-posterior (AP) view which also depict the placement of devices such as\ncentral vascular lines and tubes. In this paper, we present a new chest X-ray\nbenchmark database of 73 rich sentence-level descriptors of findings seen in AP\nchest X-rays. We describe our method of obtaining these findings through a\nsemi-automated ground truth generation process from crowdsourcing of clinician\nannotations. We also present results of building classifiers for these findings\nthat show that such higher granularity labels can also be learned through the\nframework of deep learning classifiers.\n