Blog
About

0
views
0
recommends
+1 Recommend
1 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Unveiling COVID-19 from CHEST X-Ray with Deep Learning: A Hurdles Race with Small Data

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          The possibility to use widespread and simple chest X-ray (CXR) imaging for early screening of COVID-19 patients is attracting much interest from both the clinical and the AI community. In this study we provide insights and also raise warnings on what is reasonable to expect by applying deep learning to COVID classification of CXR images. We provide a methodological guide and critical reading of an extensive set of statistical results that can be obtained using currently available datasets. In particular, we take the challenge posed by current small size COVID data and show how significant can be the bias introduced by transfer-learning using larger public non-COVID CXR datasets. We also contribute by providing results on a medium size COVID CXR dataset, just collected by one of the major emergency hospitals in Northern Italy during the peak of the COVID pandemic. These novel data allow us to contribute to validate the generalization capacity of preliminary results circulating in the scientific community. Our conclusions shed some light into the possibility to effectively discriminate COVID using CXR.

          Related collections

          Most cited references 40

          • Record: found
          • Abstract: not found
          • Conference Proceedings: not found

          Deep Residual Learning for Image Recognition

           K HE,  X. ZHANG,  S. REN (2020)
            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            Convolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?

              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: a descriptive study

              Summary Background A cluster of patients with coronavirus disease 2019 (COVID-19) pneumonia caused by infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were successively reported in Wuhan, China. We aimed to describe the CT findings across different timepoints throughout the disease course. Methods Patients with COVID-19 pneumonia (confirmed by next-generation sequencing or RT-PCR) who were admitted to one of two hospitals in Wuhan and who underwent serial chest CT scans were retrospectively enrolled. Patients were grouped on the basis of the interval between symptom onset and the first CT scan: group 1 (subclinical patients; scans done before symptom onset), group 2 (scans done ≤1 week after symptom onset), group 3 (>1 week to 2 weeks), and group 4 (>2 weeks to 3 weeks). Imaging features and their distribution were analysed and compared across the four groups. Findings 81 patients admitted to hospital between Dec 20, 2019, and Jan 23, 2020, were retrospectively enrolled. The cohort included 42 (52%) men and 39 (48%) women, and the mean age was 49·5 years (SD 11·0). The mean number of involved lung segments was 10·5 (SD 6·4) overall, 2·8 (3·3) in group 1, 11·1 (5·4) in group 2, 13·0 (5·7) in group 3, and 12·1 (5·9) in group 4. The predominant pattern of abnormality observed was bilateral (64 [79%] patients), peripheral (44 [54%]), ill-defined (66 [81%]), and ground-glass opacification (53 [65%]), mainly involving the right lower lobes (225 [27%] of 849 affected segments). In group 1 (n=15), the predominant pattern was unilateral (nine [60%]) and multifocal (eight [53%]) ground-glass opacities (14 [93%]). Lesions quickly evolved to bilateral (19 [90%]), diffuse (11 [52%]) ground-glass opacity predominance (17 [81%]) in group 2 (n=21). Thereafter, the prevalence of ground-glass opacities continued to decrease (17 [57%] of 30 patients in group 3, and five [33%] of 15 in group 4), and consolidation and mixed patterns became more frequent (12 [40%] in group 3, eight [53%] in group 4). Interpretation COVID-19 pneumonia manifests with chest CT imaging abnormalities, even in asymptomatic patients, with rapid evolution from focal unilateral to diffuse bilateral ground-glass opacities that progressed to or co-existed with consolidations within 1–3 weeks. Combining assessment of imaging features with clinical and laboratory findings could facilitate early diagnosis of COVID-19 pneumonia. Funding None.
                Bookmark

                Author and article information

                Journal
                Int J Environ Res Public Health
                Int J Environ Res Public Health
                ijerph
                International Journal of Environmental Research and Public Health
                MDPI
                1661-7827
                1660-4601
                22 September 2020
                September 2020
                : 17
                : 18
                Affiliations
                [1 ]Computer Science Department, University of Turin, 10149 Torino, Italy; carlo.barbano@ 123456edu.unito.it (C.A.B.); marco.grangetto@ 123456unito.it (M.G.)
                [2 ]Azienda Ospedaliera Città della Salute e della Scienza Presidio Molinette, 10126 Torino, Italy; cberzovini@ 123456cittadellasalute.to.it
                [3 ]Oncology Department, University of Turin, AOU San Luigi Gonzaga, 10043 Orbassano, Italy; marco.calandri@ 123456unito.it
                Author notes
                Article
                ijerph-17-06933
                10.3390/ijerph17186933
                7557723
                32971995
                © 2020 by the authors.

                Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ( http://creativecommons.org/licenses/by/4.0/).

                Categories
                Article

                Public health

                covid-19, chest x-ray, deep learning, classification

                Comments

                Comment on this article