Comparing machine learning algorithms to predict COVID‑19 mortality using a dataset including chest computed tomography severity score data

Abstract

Since the beginning of the COVID-19 pandemic, new and non-invasive digital technologies such as artificial intelligence (AI) had been introduced for mortality prediction of COVID-19 patients. The prognostic performances of the machine learning (ML)-based models for predicting clinical outcomes of COVID-19 patients had been mainly evaluated using demographics, risk factors, clinical manifestations, and laboratory results. There is a lack of information about the prognostic role of imaging manifestations in combination with demographics, clinical manifestations, and laboratory predictors. The purpose of the present study is to develop an efficient ML prognostic model based on a more comprehensive dataset including chest CT severity score (CT-SS). Fifty-five primary features in six main classes were retrospectively reviewed for 6854 suspected cases. The independence test of Chi-square was used to determine the most important features in the mortality prediction of COVID-19 patients. The most relevant predictors were used to train and test ML algorithms. The predictive models were developed using eight ML algorithms including the J48 decision tree (J48), support vector machine (SVM), multi-layer perceptron (MLP), k-nearest neighbourhood (k-NN), Naïve Bayes (NB), logistic regression (LR), random forest (RF), and eXtreme gradient boosting (XGBoost). The performances of the predictive models were evaluated using accuracy, precision, sensitivity, specificity, and area under the ROC curve (AUC) metrics. After applying the exclusion criteria, a total of 815 positive RT-PCR patients were the final sample size, where 54.85% of the patients were male and the mean age of the study population was 57.22 ± 16.76 years. The RF algorithm with an accuracy of 97.2%, the sensitivity of 100%, a precision of 94.8%, specificity of 94.5%, F1-score of 97.3%, and AUC of 99.9% had the best performance. Other ML algorithms with AUC ranging from 81.2 to 93.9% had also good prediction performances in predicting COVID-19 mortality. Results showed that timely and accurate risk stratification of COVID-19 patients could be performed using ML-based predictive models fed by routine data. The proposed algorithm with the more comprehensive dataset including CT-SS could efficiently predict the mortality of COVID-19 patients. This could lead to promptly targeting high-risk patients on admission, the optimal use of hospital resources, and an increased probability of survival of patients.

Automatic COVID-19 Prediction with Comprehensible Machine Learning Models

“…Shahriare Satu et al [26], the suggested technique may precisely predict the daily count of infected patients by training it using sample data from our web application's 25 days of data collection. Based on a larger dataset that includes the chest CT severity score (CT-SS), Seyed Zakariaee et al [27] proposed an efficient machine learning prognostic model. There were 55 primary features in 6 main classes based on retrospective reviews of 6854 probable cases.…”

Use of machine learning to identify protective factors for death from COVID-19 in the ICU: a retrospective study

“…Investigating the variables correlated with the death of patients with COVID-19 in the ICU may be fundamental for understanding the disease and improving treatment strategies ( Liu et al, 2020 ; Pijls et al, 2021 ). Data of patients with COVID-19 has been explored utilizing ML to forecast their prognosis ( Kamel et al, 2023 ) because it represents a sophisticated and adaptable approach to classification modeling by analyzing large datasets to unveil significant latent relationships or patterns ( Zakariaee et al, 2023 ). Studies indicate that, in predicting clinical outcomes among COVID-19 patients, ML methods demonstrate superior accuracy compared to conventional statistical models ( Afrash et al, 2022 ).…”

Machine learning algorithms for predicting COVID-19 mortality in Ethiopia

“…Various studies have utilized different machine-learning algorithms to identify features that can be used to predict mortality in COVID-19 patients. These features include age [ 12 , 17 , 37 – 45 ], gender [ 11 , 18 , 28 , 37 , 39 , 40 , 43 – 46 ], dry cough [ 15 , 17 , 18 , 28 , 37 , 40 , 41 , 43 , 47 ], as the clinical symptom, underlying diseases including cardiovascular disease [ 37 , 38 , 40 – 42 , 46 , 48 , 49 ], hypertension [ 37 , 38 , 41 , 43 , 44 , 46 , 50 ], diabetes [ 37 – 40 ], neurological disease [ 37 , 39 , 40 ], cancer [ 12 , 37 , 40 , 43 , 49 ]. Additionally laboratory indices such as serum creatinine [ 37 , 40 ], RBC [ 37 ], WBC [ 15 , 37 , 43 ], hematocrit [ 37 ], absolute lymphocyte count [ 11 , 28 , 37 , 40 , 41 , 46 , 47 ], absolute neutrophil count [ 15 , 17 , 28 , 37 , 40 – 42 , 47 , ...…”

“…However, these studies did not take into account important factors such as demographic, clinical, and laboratory predictors of COVID-19 mortality. It has been observed that these features have a correlation with the mortality of individuals during hospitalization [ 40 , 41 , 45 ]. To address this gap, new non-invasive digital technologies, including machine-learning prediction have been introduced for predicting the mortality of COVID-19 patients.…”