Machine learning automatically detects COVID-19 using chest CTs in a large multicenter cohort

Abstract

Objectives To investigate machine learning classifiers and interpretable models using chest CT for detection of COVID-19 and differentiation from other pneumonias, interstitial lung disease (ILD) and normal CTs. Methods Our retrospective multi-institutional study obtained 2446 chest CTs from 16 institutions (including 1161 COVID-19 patients). Training/validation/testing cohorts included 1011/50/100 COVID-19, 388/16/33 ILD, 189/16/33 other pneumonias, and 559/17/34 normal (no pathologies) CTs. A metric-based approach for the classification of COVID-19 used interpretable features, relying on logistic regression and random forests. A deep learning–based classifier differentiated COVID-19 via 3D features extracted directly from CT attenuation and probability distribution of airspace opacities. Results Most discriminative features of COVID-19 are the percentage of airspace opacity and peripheral and basal predominant opacities, concordant with the typical characterization of COVID-19 in the literature. Unsupervised hierarchical clustering compares feature distribution across COVID-19 and control cohorts. The metrics-based classifier achieved AUC = 0.83, sensitivity = 0.74, and specificity = 0.79 versus respectively 0.93, 0.90, and 0.83 for the DL-based classifier. Most of ambiguity comes from non-COVID-19 pneumonia with manifestations that overlap with COVID-19, as well as mild COVID-19 cases. Non-COVID-19 classification performance is 91% for ILD, 64% for other pneumonias, and 94% for no pathologies, which demonstrates the robustness of our method against different compositions of control groups. Conclusions Our new method accurately discriminates COVID-19 from other types of pneumonia, ILD, and CTs with no pathologies, using quantitative imaging features derived from chest CT, while balancing interpretability of results and classification performance and, therefore, may be useful to facilitate diagnosis of COVID-19. Key Points • Unsupervised clustering reveals the key tomographic features including percent airspace opacity and peripheral and basal opacities most typical of COVID-19 relative to control groups. • COVID-19-positive CTs were compared with COVID-19-negative chest CTs (including a balanced distribution of non-COVID-19 pneumonia, ILD, and no pathologies). Classification accuracies for COVID-19, pneumonia, ILD, and CT scans with no pathologies are respectively 90%, 64%, 91%, and 94%. • Our deep learning (DL)–based classification method demonstrates an AUC of 0.93 (sensitivity 90%, specificity 83%). Machine learning methods applied to quantitative chest CT metrics can therefore improve diagnostic accuracy in suspected COVID-19, particularly in resource-constrained environments. Supplementary Information The online version contains supplementary material available at 10.1007/s00330-021-07937-3.

A Promising and Challenging Approach: Radiologists’ Perspective on Deep Learning and Artificial Intelligence for Fighting COVID-19

“…When detecting COVID-19 and CAP, the area under the curve (AUC) of the receiver operating characteristic (ROC) was 0.96 and 0.95, respectively. The study of Barbosa et al achieved similar results [11]. This research shows that the DL and AI models trained by collecting large numbers of patient image datasets can efficiently screen and diagnose suspected COVID-19 patients.…”

Detection and characterization of COVID-19 findings in chest CT

“…Technical details for the evolution of the algorithm have been described before. [17,23] It provides opacity scores, percentages of opacity (relative to overall lung volume), and percentages of high opacity (relative to overall lung volume). To distinguish between ground glass opacities and consolidations, a threshold of À200 HU is applied inside the detected airspace opacities.…”

“…The algorithm was trained using international multicentre datasets. [17] The primary objective of this study was to evaluate the feasibility and applicability of an AI-based software prototype to detect COVID associated lung abnormalities in chest CT. In clinical routine the prototype was applied in patients with known COVID-19 disease and an asymptomatic control cohort.…”

Commercial AI solutions in detecting COVID-19 pneumonia in chest CT: not yet ready for clinical implementation?

“…At the time of the initial evaluation of our chest CT collection, all solutions were under development and not purchasable. To date, the training and validations of two of the evaluated algorithms have been published [ 18 , 19 ]. Both algorithms were trained to detect chest CTs suspicious for COVID-19 pneumonia and achieved a sensitivity ≥ 0.90 and a specificity ≥ 0.83 in training.…”