An End-to-End Integrated Clinical and CT Based Radiomics Nomogram for Predicting Disease Severity and Need for Ventilator Support in COVID-19 Patients: A Large Multi-Site Retrospective Study

Abstract

Objective: The disease COVID-19 has caused a widespread global pandemic with ~3.93 million deaths worldwide. In this work, we present three models- Radiomics (MRM), Clinical (MCM), and combined Clinical-Radiomics (MRCM) nomogram to predict COVID-19 positive patients who will end up needing invasive mechanical ventilation from the baseline CT scans. Method: We performed a retrospective multicohort study of individuals with COVID-19 positive findings for a total of 980 patients from 2 different institutions (Renmin hospital of Wuhan University, D1 =787 and University Hospitals, US D2 = 110). The patients from institution-1 were divided into 60% training, D1 T (N=473), and 40% test set D1 V (N=314). The patients from institution-2 were used for an independent validation test set D2 V(N=110). A U-Net-based neural network (CNN) was trained to automatically segment out the COVID consolidation regions on the CT scans. The segmented regions from the CT scans were used for extracting first-order and higher-order Radiomic textural features. The top Radiomic and clinical features were selected using the Least Absolute Shrinkage and Selection Operator (LASSO) with an optimal binomial regression model within D1 T. Results: The 3 out of the top 5 features identified using D1 T were higher-order textural features (GLCM, GLRLM, GLSZM), whereas the last two features included the total infection size on the CT scan and the total intensity of the COVID consolidations. The Radiomics Model (MRM ) was constructed using the Radiomic Score built using the coefficients obtained from the LASSO logistic model used within the linear regression (LR) classifier. The MRM yielded an area under the receiver operating characteristic curve (AUC) of 0.754 [0.709-0.799] on D1 T, 0.836 on D1 V, and 0.748 D2 V. The top prognostic clinical factors identified in the analysis were dehydrogenase (LDH), age, and albumin (ALB). The clinical model had an AUC of 0.784 [0.743-0.825] on D1 T, 0.813 on D1 V, and 0.723 on D2 V. Finally, the combined model, MRCM integrating Radiomic Score, age, LDH and ALB, yielded an AUC of 0.814 [0.774-0.853] on D1 T, 0.847 on D1 V, and 0.772 on D2 V. The MRCM had an overall improvement in the performance of ~3.77% (D1 T: p = 0.0003; D1 V p= 0.0165; D2 V : p = 0.024) over MCM Conclusion: The novel integrated imaging and clinical model (MRCM) outperformed both models (MRM) and (MCM). Our results across multiple sites suggest that the integrated nomogram could help identify COVID-19 patients with more severe disease phenotype and potentially requiring mechanical ventilation. Funding: National Cancer Institute of the US National Institutes of Health, National Center for Research Resources, US Department of Veterans Affairs Biomedical Laboratory Research and Development Service, Department of Defense, National Institute of Diabetes and Digestive and Kidney Diseases, Wallace H Coulter Foundation, Case Western Reserve University, and Dana Foundation. Declaration of Interest: AM reports grants from National Cancer Institute of the National Institutes of Health, grants from National Center for Research Resources, grants from VA Merit Review Award, grants from DOD Cancer Investigator-Initiated Translational Research Award, during the conduct of the study; grants from DOD Prostate Cancer Idea Development Award, grants from DOD Peer Reviewed Cancer Research Program, grants from National Institute of Diabetes and Digestive and Kidney Diseases , grants from the Ohio Third Frontier Technology Validation Fund, grants from the Wallace H. Coulter Foundation Program in the Department of Biomedical Engineering and the Clinical and Translational Science Award Program (CTSA) at Case Western Reserve University, grants from Department of Defense Peer Reviewed Cancer Research Program (PRCRP) Career Development Award, grants from Dana Foundation David Mahoney Neuroimaging Program. Ethical Approval: The study conformed to HIPAA guidelines was approved by University Hospitals, Cleveland (STUDY20200463), and the Ethics committee of the Renmin Hospital of Wuhan University (ethics number: V1.0).