Biomaterial Database

A Deep-Learning Diagnostic Support System for the Detection of COVID-19 Using Chest Radiographs: A Multireader Validation Study. 2021

Matthias Fontanellaz, and Lukas Ebner, and Adrian Huber, and Alan Peters, and Laura Löbelenz, and Cynthia Hourscht, and Jeremias Klaus, and Jaro Munz, and Thomas Ruder, and Dionysios Drakopoulos, and Dominik Sieron, and Elias Primetis, and Johannes T Heverhagen, and Stavroula Mougiakakou, and Andreas Christe

From the ARTORG Center for Biomedical Engineering Research, University of Bern.

Five publicly available databases comprising normal CXR, confirmed COVID-19 pneumonia cases, and other pneumonias were used. After the harmonization of the data, the training set included 7966 normal cases, 5451 with other pneumonia, and 258 CXRs with COVID-19 pneumonia, whereas in the testing data set, each category was represented by 100 cases. Eleven blinded radiologists with various levels of expertise independently read the testing data set. The data were analyzed separately with the newly proposed artificial intelligence-based system and by consultant radiologists and residents, with respect to positive predictive value (PPV), sensitivity, and F-score (harmonic mean for PPV and sensitivity). The χ2 test was used to compare the sensitivity, specificity, accuracy, PPV, and F-scores of the readers and the system. The proposed system achieved higher overall diagnostic accuracy (94.3%) than the radiologists (61.4% ± 5.3%). The radiologists reached average sensitivities for normal CXR, other type of pneumonia, and COVID-19 pneumonia of 85.0% ± 12.8%, 60.1% ± 12.2%, and 53.2% ± 11.2%, respectively, which were significantly lower than the results achieved by the algorithm (98.0%, 88.0%, and 97.0%; P < 0.00032). The mean PPVs for all 11 radiologists for the 3 categories were 82.4%, 59.0%, and 59.0% for the healthy, other pneumonia, and COVID-19 pneumonia, respectively, resulting in an F-score of 65.5% ± 12.4%, which was significantly lower than the F-score of the algorithm (94.3% ± 2.0%, P < 0.00001). When other pneumonia and COVID-19 pneumonia cases were pooled, the proposed system reached an accuracy of 95.7% for any pathology and the radiologists, 88.8%. The overall accuracy of consultants did not vary significantly compared with residents (65.0% ± 5.8% vs 67.4% ± 4.2%); however, consultants detected significantly more COVID-19 pneumonia cases (P = 0.008) and less healthy cases (P < 0.00001). The system showed robust accuracy for COVID-19 pneumonia detection on CXR and surpassed radiologists at various training levels.

UI	MeSH Term	Description	Entries
D007091	Image Processing, Computer-Assisted	A technique of inputting two-dimensional or three-dimensional images into a computer and then enhancing or analyzing the imagery into a form that is more useful to the human observer.	Biomedical Image Processing,Computer-Assisted Image Processing,Digital Image Processing,Image Analysis, Computer-Assisted,Image Reconstruction,Medical Image Processing,Analysis, Computer-Assisted Image,Computer-Assisted Image Analysis,Computer Assisted Image Analysis,Computer Assisted Image Processing,Computer-Assisted Image Analyses,Image Analyses, Computer-Assisted,Image Analysis, Computer Assisted,Image Processing, Biomedical,Image Processing, Computer Assisted,Image Processing, Digital,Image Processing, Medical,Image Processings, Medical,Image Reconstructions,Medical Image Processings,Processing, Biomedical Image,Processing, Digital Image,Processing, Medical Image,Processings, Digital Image,Processings, Medical Image,Reconstruction, Image,Reconstructions, Image
D011237	Predictive Value of Tests	In screening and diagnostic tests, the probability that a person with a positive test is a true positive (i.e., has the disease), is referred to as the predictive value of a positive test; whereas, the predictive value of a negative test is the probability that the person with a negative test does not have the disease. Predictive value is related to the sensitivity and specificity of the test.	Negative Predictive Value,Positive Predictive Value,Predictive Value Of Test,Predictive Values Of Tests,Negative Predictive Values,Positive Predictive Values,Predictive Value, Negative,Predictive Value, Positive
D005260	Female		Females
D006801	Humans	Members of the species Homo sapiens.	Homo sapiens,Man (Taxonomy),Human,Man, Modern,Modern Man
D000077321	Deep Learning	Supervised or unsupervised machine learning methods that use multiple layers of data representations generated by nonlinear transformations, instead of individual task-specific ALGORITHMS, to build and train neural network models.	Hierarchical Learning,Learning, Deep,Learning, Hierarchical
D000086382	COVID-19	A viral disorder generally characterized by high FEVER; COUGH; DYSPNEA; CHILLS; PERSISTENT TREMOR; MUSCLE PAIN; HEADACHE; SORE THROAT; a new loss of taste and/or smell (see AGEUSIA and ANOSMIA) and other symptoms of a VIRAL PNEUMONIA. In severe cases, a myriad of coagulopathy associated symptoms often correlating with COVID-19 severity is seen (e.g., BLOOD COAGULATION; THROMBOSIS; ACUTE RESPIRATORY DISTRESS SYNDROME; SEIZURES; HEART ATTACK; STROKE; multiple CEREBRAL INFARCTIONS; KIDNEY FAILURE; catastrophic ANTIPHOSPHOLIPID ANTIBODY SYNDROME and/or DISSEMINATED INTRAVASCULAR COAGULATION). In younger patients, rare inflammatory syndromes are sometimes associated with COVID-19 (e.g., atypical KAWASAKI SYNDROME; TOXIC SHOCK SYNDROME; pediatric multisystem inflammatory disease; and CYTOKINE STORM SYNDROME). A coronavirus, SARS-CoV-2, in the genus BETACORONAVIRUS is the causative agent.	2019 Novel Coronavirus Disease,2019 Novel Coronavirus Infection,2019-nCoV Disease,2019-nCoV Infection,COVID-19 Pandemic,COVID-19 Pandemics,COVID-19 Virus Disease,COVID-19 Virus Infection,Coronavirus Disease 2019,Coronavirus Disease-19,SARS Coronavirus 2 Infection,SARS-CoV-2 Infection,Severe Acute Respiratory Syndrome Coronavirus 2 Infection,COVID19,2019 nCoV Disease,2019 nCoV Infection,2019-nCoV Diseases,2019-nCoV Infections,COVID 19,COVID 19 Pandemic,COVID 19 Virus Disease,COVID 19 Virus Infection,COVID-19 Virus Diseases,COVID-19 Virus Infections,Coronavirus Disease 19,Disease 2019, Coronavirus,Disease, 2019-nCoV,Disease, COVID-19 Virus,Infection, 2019-nCoV,Infection, COVID-19 Virus,Infection, SARS-CoV-2,Pandemic, COVID-19,SARS CoV 2 Infection,SARS-CoV-2 Infections,Virus Disease, COVID-19,Virus Infection, COVID-19
D012189	Retrospective Studies	Studies used to test etiologic hypotheses in which inferences about an exposure to putative causal factors are derived from data relating to characteristics of persons under study or to events or experiences in their past. The essential feature is that some of the persons under study have the disease or outcome of interest and their characteristics are compared with those of unaffected persons.	Retrospective Study,Studies, Retrospective,Study, Retrospective
D013902	Radiography, Thoracic	X-ray visualization of the chest and organs of the thoracic cavity. It is not restricted to visualization of the lungs.	Thoracic Radiography,Radiographies, Thoracic,Thoracic Radiographies