Deep learning-based lesion subtyping and prediction of clinical outcomes in COVID-19 pneumonia using chest CT

Abstract

The main objective of this work is to develop and evaluate an artificial intelligence system based on deep learning capable of automatically identifying, quantifying, and characterizing COVID-19 pneumonia patterns in order to assess disease severity and predict clinical outcomes, and to compare the prediction performance with respect to human reader severity assessment and whole lung radiomics. We propose a deep learning based scheme to automatically segment the different lesion subtypes in nonenhanced CT scans. The automatic lesion quantification was used to predict clinical outcomes. The proposed technique has been independently tested in a multicentric cohort of 103 patients, retrospectively collected between March and July of 2020. Segmentation of lesion subtypes was evaluated using both overlapping (Dice) and distance-based (Hausdorff and average surface) metrics, while the proposed system to predict clinically relevant outcomes was assessed using the area under the curve (AUC). Additionally, other metrics including sensitivity, specificity, positive predictive value and negative predictive value were estimated. 95% confidence intervals were properly calculated. The agreement between the automatic estimate of parenchymal damage (%) and the radiologists’ severity scoring was strong, with a Spearman correlation coefficient (R) of 0.83. The automatic quantification of lesion subtypes was able to predict patient mortality, admission to the Intensive Care Units (ICU) and need for mechanical ventilation with an AUC of 0.87, 0.73 and 0.68 respectively. The proposed artificial intelligence system enabled a better prediction of those clinically relevant outcomes when compared to the radiologists’ interpretation and to whole lung radiomics. In conclusion, deep learning lesion subtyping in COVID-19 pneumonia from noncontrast chest CT enables quantitative assessment of disease severity and better prediction of clinical outcomes with respect to whole lung radiomics or radiologists’ severity score.

Prediction of prognosis in patients with severe COVID-19 pneumonia using CT score by emergency physicians: a single-center retrospective study

“…COVID is a pandemic, and it is not always possible to treat patients in hospitals equipped with such facilities. A complementary method is a deep-learning reading system, the usefulness of which has been previously reported 24 . We conducted this study in the hopes that our generated CT score could be a tool that could help in situations where it was not possible to prepare such a score.…”

A Multiclass Radiomics Method–Based WHO Severity Scale for Improving COVID-19 Patient Assessment and Disease Characterization From CT Scans

“…On the other hand, other studies uses radiomics features extracted from medical images to predict disease outcomes and used the radiomics features to characterize COVID-19 severity. [36][37][38] We advanced prior studies by categorizing COVID-19 severity into 4-disease statuses using the WHO-CPS. We also characterized COVID-19 lesions in 4-lung pathologies and created a "severity signature" by utilizing radiomics characteristics from each lung pathology.…”

“…In addition, we automated the multiclass radiomics model AssessNet-19 to address CT quantification, severity assessment, and disease characterization. On the other hand, other studies uses radiomics features extracted from medical images to predict disease outcomes and used the radiomics features to characterize COVID-19 severity 36–38 . We advanced prior studies by categorizing COVID-19 severity into 4-disease statuses using the WHO-CPS.…”

Improve the performance of CT-based pneumonia classification via source data reweighting

“…Chaudhary et al 46 developed a two-stage convolutional neural network (CNN) to detect COVID-19 and community acquired pneumonia (CAP) from CT scans. Bermejo-Peláez et al 47 proposed a deep neural network to analyze COVID-19 patterns from CT scans to assess disease severity and predict clinical outcomes. Yao et al 48 developed an atrous convolution network to diagnose mild COVID-19 pneumonia from CT scans.…”