Juan Carlos Quiroz1,2, You-Zhen Feng3, Zhong-Yuan Cheng3, Dana Rezazadegan1,4, Ping-Kang Chen3, Qi-Ting Lin3, Long Qian5, Xiao-Fang Liu6,7, Shlomo Berkovsky1, Enrico Coiera1, Lei Song8, Xiao-Ming Qiu9, Sidong Liu1, Xiang-Ran Cai3. 1. Centre for Health Informatics, Australian Institute of Health Innovation, Faculty of Medicine, Health and Human Sciences, Macquarie University, Macquarie University75 Talavera Road, Macquarie Park, AU. 2. Centre for Big Data Research in Health, University of New South Wales, Sydney, AU. 3. Medical Imaging Centre, The First Affiliated Hospital of Jinan University, Guangzhou, CN. 4. Department of Computer Science and Software Engineering, Swinburne University of Technology, Melbourne, AU. 5. Department of Biomedical Engineering, Peking University, Beijing, CN. 6. Institute of Robotics and Automatic Information System, College of Artificial Intelligence, Nankai University, Tianjin, CN. 7. School of Data and Computer Science, Sun Yat-sen University, Guangzhou, CN. 8. Department of Radiology, Xiangyang Central Hospital, Affiliated Hospital of Hubei University of Arts and Science, Xiangyang, CN. 9. Department of Radiology, Huangshi Central Hospital, Affiliated Hospital of Hubei Polytechnic University, Edong Healthcare Group, Huangshi, CN.
Abstract
BACKGROUND: Coronavirus disease 2019 (COVID-19) has overwhelmed health systems worldwide. It is important to identify severe cases as early as possible, so that resources can be mobilized and treatment can be escalated. OBJECTIVE: This study aims to develop a machine learning approach for automated severity assessment of COVID-19 patients based on clinical and imaging data. METHODS: Clinical data-demographics, signs, symptoms, comorbidities and blood test results-and chest computer tomography (CT) scans of 346 patients from two hospitals in the Hubei province, China, were used to develop machine learning models for automated severity assessment of diagnosed COVID-19 cases. We compared the predictive power of clinical and imaging data by testing multiple machine learning models, and further explored the use of four oversampling methods to address the imbalance distribution issue. Features with the highest predictive power were identified using the SHapley Additive exPlanations (SHAP) framework. RESULTS: Imaging features had the strongest impact on the model output, while a combination of clinical and imaging features yielded the best performance overall. The identified predictive features were consistent with findings from previous studies. Oversampling yielded mixed results, although it achieved the best model performance in our study. Targeting differentiation between mild and severe cases, logistic regression models achieved the best performance on clinical features (area under the curve [AUC]:0.848, sensitivity:0.455, specificity:0.906), imaging features (AUC:0.926, sensitivity:0.818, specificity:0.901) and the combined features (AUC:0.950, sensitivity:0.764, specificity:0.919). The SMOTE oversampling method further improved the performance of the combined features to AUC of 0.960 (sensitivity:0.845, specificity:0.929). CONCLUSIONS: This study indicates that clinical and imaging features can be used for automated severity assessment of COVID-19 patients and have the potential to assist with triaging COVID-19 patients and prioritizing care for patients at higher risk of severe cases.
BACKGROUND:Coronavirus disease 2019 (COVID-19) has overwhelmed health systems worldwide. It is important to identify severe cases as early as possible, so that resources can be mobilized and treatment can be escalated. OBJECTIVE: This study aims to develop a machine learning approach for automated severity assessment of COVID-19patients based on clinical and imaging data. METHODS: Clinical data-demographics, signs, symptoms, comorbidities and blood test results-and chest computer tomography (CT) scans of 346 patients from two hospitals in the Hubei province, China, were used to develop machine learning models for automated severity assessment of diagnosed COVID-19 cases. We compared the predictive power of clinical and imaging data by testing multiple machine learning models, and further explored the use of four oversampling methods to address the imbalance distribution issue. Features with the highest predictive power were identified using the SHapley Additive exPlanations (SHAP) framework. RESULTS: Imaging features had the strongest impact on the model output, while a combination of clinical and imaging features yielded the best performance overall. The identified predictive features were consistent with findings from previous studies. Oversampling yielded mixed results, although it achieved the best model performance in our study. Targeting differentiation between mild and severe cases, logistic regression models achieved the best performance on clinical features (area under the curve [AUC]:0.848, sensitivity:0.455, specificity:0.906), imaging features (AUC:0.926, sensitivity:0.818, specificity:0.901) and the combined features (AUC:0.950, sensitivity:0.764, specificity:0.919). The SMOTE oversampling method further improved the performance of the combined features to AUC of 0.960 (sensitivity:0.845, specificity:0.929). CONCLUSIONS: This study indicates that clinical and imaging features can be used for automated severity assessment of COVID-19patients and have the potential to assist with triaging COVID-19patients and prioritizing care for patients at higher risk of severe cases.
Authors: Karrar Hameed Abdulkareem; Ammar Awad Mutlag; Ahmed Musa Dinar; Jaroslav Frnda; Mazin Abed Mohammed; Fawzi Hasan Zayr; Abdullah Lakhan; Seifedine Kadry; Hasan Ali Khattak; Jan Nedoma Journal: Comput Intell Neurosci Date: 2022-07-19
Authors: Liam Butler; Ibrahim Karabayir; Mohammad Samie Tootooni; Majid Afshar; Ari Goldberg; Oguz Akbilgic Journal: Int J Med Inform Date: 2021-12-09 Impact factor: 4.730
Authors: Karrar Hameed Abdulkareem; Mohammed Nasser Al-Mhiqani; Ahmed M Dinar; Mazin Abed Mohammed; Mustafa Jawad Al-Imari; Alaa S Al-Waisy; Abed Saif Alghawli; Mohammed A A Al-Qaness Journal: Bioengineering (Basel) Date: 2022-09-08
Authors: Grigoris T Gerotziafas; Mariella Catalano; Yiannis Theodorou; Patrick Van Dreden; Vincent Marechal; Alex C Spyropoulos; Charles Carter; Nusrat Jabeen; Job Harenberg; Ismail Elalamy; Anna Falanga; Jawed Fareed; Petros Agathaggelou; Darko Antic; Pier Luigi Antignani; Manuel Monreal Bosch; Benjamin Brenner; Vladimir Chekhonin; Mary-Paula Colgan; Meletios-Athanasios Dimopoulos; Jim Douketis; Essam Abo Elnazar; Katalin Farkas; Bahare Fazeli; Gerry Fowkes; Yongquan Gu; Joseph Gligorov; Mark A Ligocki; Tishya Indran; Meganathan Kannan; Bulent Kantarcioglu; Abdoul Aziz Kasse; Kostantinos Konstantinidis; Fabio Leivano; Joseph Lewis; Alexander Makatsariya; P Massamba Mbaye; Isabelle Mahé; Irina Panovska-Stavridis; Dan-Mircea Olinic; Chryssa Papageorgiou; Zsolt Pecsvarady; Sergio Pillon; Eduardo Ramacciotti; Hikmat Abdel-Razeq; Michele Sabbah; Mouna Sassi; Gerit Schernthaner; Fakiha Siddiqui; Jin Shiomura; Anny Slama-Schwok; Jean Claude Wautrecht; Alfonso Tafur; Ali Taher; Peter Klein-Wegel; Zenguo Zhai; Tazi Mezalek Zoubida Journal: Thromb Haemost Date: 2021-07-20 Impact factor: 6.681