Skip to main content

Prognostic significance of chest CT severity score in mortality prediction of COVID-19 patients, a machine learning study

Abstract

Background

The high mortality rate of COVID-19 makes it necessary to seek early identification of high-risk patients with poor prognoses. Although the association between CT-SS and mortality of COVID-19 patients was reported, its prognosis significance in combination with other prognostic parameters was not evaluated yet.

Methods

This retrospective single-center study reviewed a total of 6854 suspected patients referred to Imam Khomeini hospital, Ilam city, west of Iran, from February 9, 2020 to December 20, 2020. The prognostic performances of k-Nearest Neighbors (kNN), Multilayer Perceptron (MLP), Support Vector Machine (SVM), and J48 decision tree algorithms were evaluated based on the most important and relevant predictors. The metrics derived from the confusion matrix were used to determine the performance of the ML models.

Results

After applying exclusion criteria, 815 hospitalized cases were entered into the study. Of these, 447(54.85%) were male and the mean (± SD) age of participants was 57.22(± 16.76) years. The results showed that the performances of the ML algorithms were improved when they are fed by the dataset with CT-SS data. The kNN model with an accuracy of 94.1%, sensitivity of 100. 0%, precision of 89.5%, specificity of 88.3%, and AUC around 97.2% had the best performance among the other three ML techniques.

Conclusions

The integration of CT-SS data with demographics, risk factors, clinical manifestations, and laboratory parameters improved the prognostic performances of the ML algorithms. An ML model with a comprehensive collection of predictors could identify high-risk patients more efficiently and lead to the optimal use of hospital resources.

Background

Coronavirus disease 2019 (COVID-19) is a life-threatening infection caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1]. Despite all the preventive and lockdown measures taken by governments, the COVID-19 outbreak continues to spread aggressively worldwide and an exponential daily increasing number of infected cases is reported. The complex and highly contagious nature of COVID-19 has made this infection a serious global health concern and a notable pandemic [2,3,4,5]. The clinical outcomes of COVID-19 range from asymptomatic, mild or moderate symptoms to serious complications and death in some cases [6].

This virus has made a tremendous impact on the health status of people all over the world and caused a significant number of deaths. Approximately 20% of COVID-19 patients need to be hospitalized [7] and the pooled case fatality rate (CFR) of these patients is 13%. While, CFR in patients admitted in intensive care unit (ICU) would be 37.0% [8]. The high mortality rate of COVID-19 particularly for elderly populations and patients with underlying comorbidities including cardiopulmonary diseases, cancer, hypertension, diabetes, and low immune functions makes it necessary to seek early identification of high-risk patients with poor prognoses. Due to the unpredictability of the disease behavior and courses, the prognosis of disease progression in hospitalized patients and the identification of patients prone to rapid deterioration is a challenging clinical problem. Clinicians and health policymakers have commonly used predictions made by different statistical models to deal with these challenges [9, 10]. As a good alternative, artificial intelligence (AI) may be a helpful tool to identify patients with high-risk of mortality. AI is a noninvasive digital technology that can facilitate accurate and timely identification of high-risk patients. Machine learning (ML) is a subset of AI that looks for hidden and previously unknown patterns from large sets of data [11]. In the prior studies, ML-based models were evaluated to predict the risk of patient deterioration and death. These studies mainly used demographics, risk factors, clinical manifestations, and laboratory results [12,13,14,15]. In recent meta-analysis studies, it was shown that chest computed tomography severity score (CT-SS) is an appropriate prognostic factor for mortality prediction in COVID-19 patients [16, 17]. Therefore, it might improve the prognostic performances of the ML algorithms for predicting clinical outcomes of patients with COVID-19 pneumonia. In this study, we evaluated the prognostic significance of CT-SS in the mortality prediction of COVID-19 patients using the selected well-known ML algorithms. The prognostic performances of four ML algorithms including kNN, MLP, SVM, and J48 decision tree algorithms are assessed in the presence and absence of CT-SS data.

Methods

Dataset description

This retrospective single-center study was conducted in 2022 to predict mortality in COVID-19 patients based on four popular ML algorithms. This study reviewed a hospital-based COVID-19 registry database from Imam Khomeini hospital, Ilam city, west of Iran, from February 9, 2020 to December 20, 2020. During this period, a total of 6854 suspected cases were referred to Imam Khomeini Hospital’s ambulatory and emergency departments (EDs), of whom 1853 cases were introduced as positive RT-PCR COVID-19, 2472 as negative, and 2529 as unknown. Only the patients with positive RT-PCR tests were included in the study (Fig. 1).

Fig. 1
figure 1

Flowchart describing patient selection

The included cases were linked to 55 primary features in six main classes including patient’s demographics (five features), clinical pictures (14 features), history of personal diseases/comorbidity (seven features), laboratory results (26 features), CT-SS (one feature), and an output variable (0: survived and 1: deceased) (see Table 1).

Table 1 The most important variable at P < 0.01 using the Chi-squared test

The severity of the pulmonary involvement for each patient was evaluated using the chest CT score. Five lung lobes were visually scored as 0 (no involvement), 1 (less than 5% involvement), 2 (5%–25% involvement), 3 (26%–49% involvement), 4 (50%– 75% involvement), and 5 (50%– 75% involvement). The sum of these scores yielded the total CT-SS ranging from 0 to 25. Two radiologists separately reviewed all CT images. Any disagreements were resolved through consulting with an attending radiologist with 23 years of experience.

Data pre-processing

In the data mining process, the use of raw data reduces the efficiencies of the algorithms and the achieved results would have poor qualities. The refined information extracted from the raw data considerably improves the model’s ability to learn. Therefore, the data pre-processing would be a crucial step before the training of the model. This approach resolves inconsistencies and addresses irrelevant, redundant, and unreliable data [18].

In this study, the incomplete records with many missing values (more than 70%) were excluded from the dataset. Noisy and abnormal values, errors, and meaningless data were checked by two authors (MB and SSZ). The missing cells of continuous and discrete variables were imputed by mean and mode values, respectively.

The schematic of the study inclusion and exclusion criteria is shown in Fig. 1. The final sample size was 815 COVID-19 patients. The refined dataset is significantly imbalanced in terms of the number of records in outcome classes. It contains 707 and 108 cases in the alive and death classes, respectively. This problem would cause delivering results biased toward the dominant class. In this study, the synthetic minority over-sampling technique (SMOTE) method was used to deal with the unbalanced dataset (https://imbalanced-learn.org/stable/).

Feature selection

The feature selection process is a beneficial statistical method commonly used in forecasting, pattern recognition, and classification modeling for determining the most important variables highly correlated with the target variable [19].

This statistical approach prevents overfitting of the data mining algorithms, results in better classification of the data and evaluation of fewer numbers of variables for work simplification [20]. In this study, the independence test of Chi-square was used for weighting the features based on their importance in the mortality prediction of COVID-19 patients. The Chi-square coefficient was calculated by the following equation:

$$x^{2} = \frac{{\left( {O_{i} - E_{i} } \right)^{2} }}{{E_{i} }}$$
(1)

where Oi and Ei are the observed and expected variables, respectively. The SPSS software (version 23) was used to determine the importance of the variables for mortality prediction of COVID-19 patients. In this study, P < 0.01 was regarded as the significant level.

Model development

In this study, four well-known ML classification methods including kNN, MLP, SVM, and J48 decision tree algorithms were used to predict the mortality of the patient with confirmed COVID-19. These algorithms were implemented using Waikato Environment for Knowledge Analysis (Weka) software (version 3.9.2, University of Waikato, New Zealand). In the performance evaluation of the developed classifiers, tenfold cross-validation method was used. In this approach, the data set would be divided into ten subsets and all models run ten times. Each time, one subset was considered as test data and the remaining nine subsets would be training datasets. The results of these ten evaluations are mixed to render the performance metrics. Therefore, tenfold cross-validation method would be a preferred technique due to its relatively low-level bias and variation. The performances of these classification algorithms were determined in terms of accuracy, precision, sensitivity, specificity, F-measure, and area under the ROC curve (AUC).

ML algorithms were applied to datasets with and without CT-SS data to determine the prognostic significance of CT-SS in the mortality prediction of COVID-19 patients.

Ethical considerations

This study was approved by the medical ethical committee (approved number: IR.MEDILAM.REC.1401.255). The unique identifying information of the patients was concealed to protect the privacy and confidentiality of the patients.

Results

A total of 6854 suspected patients was registered in the Ilam CoV registry database. After applying exclusion criteria including negative RT-PCR COVID-19 test, age lower than 18 years old, discharged or death from the emergency department, missing data more than 70%, noisy and abnormal values, unknown dispositions, and lack of laboratory or CT-SS data, 815 hospitalized cases were entered into the study. Of these, 447(54.85%) were male and the mean (± SD) age of participants was 57.22(± 16.76) years. Out of 815 included patients, 707 patients recovered and 108 (13.3%) deceased. The number of records in the deceased class was raised to 707 after balancing the dataset.

Feature selection

After utilizing the independence test of Chi-square to determine the importance of the variables for the mortality prediction of COVID-19 patients, 27 predictors were chosen as the most important and relevant features. The list of the most important variables and results of the independence test of Chi-square are demonstrated in Table 1. These features were used as the inputs for all ML algorithms. These features included demographics, risk factors, clinical manifestations, laboratory tests, and imaging results.

Evaluation of the developed models

In this study, the predictive models were built using four well-known ML algorithms including kNN, MLP, SVM, and J48 decision tree. The subsets of features selected using the independence test of Chi-square were used to develop COVID-19 mortality prediction models. The ML algorithms were separately trained using the datasets with and without CT-SS data. The performances of these models were evaluated using sensitivity, specificity, accuracy, precision, F-measure, and AUC metrics. The results of the performance evaluation for the developed models using the datasets with and without CT-SS data are listed in Table 2. The results showed that the kNN algorithm yielded better performance to predict the mortality of COVID-19 patients than other ML algorithms. Sensitivity, specificity, accuracy, precision, F-Measure, and AUC of the kNN algorithm fed by the dataset without CT-SS data were 100.0%, 87.0%, 93.5%, 88.5%, 93.9%, and 97.5%, respectively. For the dataset with CT-SS data, the kNN algorithm reached 100.0% sensitivity, 88.3% specificity, 94.1% accuracy, 89.5% precision, 94.5% F-Measure, and an AUC of 97.2%.

Table 2 Performances of ML algorithms fed by the datasets with and without CT-SS data

Figure 2 depicts the ROC curves for the selected ML algorithms fed by the datasets with and without CT-SS data, separately.

Fig. 2
figure 2

ROC curves for ML algorithms fed by the datasets with and without CT-SS data

The results showed that the performances of the algorithms were improved when they are fed by the dataset with CT-SS data.

Discussion

During the major outbreak of COVID-19, a timely and accurate prognosis of disease progression and clinical outcomes of patients could provide better guidelines for the management of the disease. An efficient predictive model to identify high-risk patients could have a crucial role in risk stratification for the allocation of finite resources and improved patient survival probability. For appropriate preparedness against this global pandemic, new technologies and AI-based solutions have been suggested for diagnostic, screening, monitoring, and prognostic goals. In these studies, various models were evaluated and the most relevant predictive parameters were reported [21,22,23,24,25,26,27].

CT-SS is a promising prognostic factor for mortality prediction in COVID-19 patients that evaluate the severity and extent of pulmonary involvement. A recent meta-analysis study showed that CT-SS index with and without an optimal cutoff was positively associated with mortality of COVID-19 patients (OR 7.124; 95% CI 5.307–9.563 and OR 1.244; 95% CI 1.157–1.337, respectively) [17]. Sensitivity, specificity, and AUC of this predictive parameter were 0.67 (95%CI: 0.59–0.75), 0.79 (95%CI: 0.74–0.84), and 0.8248, respectively [16]. Although the association between CT-SS and mortality of COVID-19 patients was reported [16, 17], its prognosis significance in combination with other prognostic parameters was not evaluated yet. Therefore, in this study, kNN, MLP, SVM, and J48 decision tree algorithms were developed based on the most relevant features in determining the risk of COVID-19 mortality. The prognostic importance of the parameters in the mortality prediction of COVID-19 patients was determined using the independence test of Chi-square. Some features such as age, cough, underlying disease, serum creatinine, and CT-SS were of the highest importance. Relevant predictors of deterioration and mortality risks of COVID-19 patients were also reported by several studies [9, 10, 12, 13, 28,29,30,31,32,33,34,35,36,37,38,39]. The most relevant predictive features reported by these studies are similar to our findings. On the other hand, some features such as smoking, alcohol consumption, drug addiction, platelet count, and alanine aminotransferase (ALT) had a low weight of significance in predicting COVID-19 mortality. It should be noted that these variables with the lowest significance in predicting COVID-19 mortality are consistent with those reported by the study of Moulaei et al. [15]. Although these factors have a considerable clinical role in the success of treatment, many of these features can be ignored in the ML approach and poor outcomes of the infected patients could be efficiently predicted with fewer factors.

In the next step, the prognostic performances of the selected ML models were evaluated based on a dataset of laboratory-confirmed COVID-19 hospitalized patients with and without CT-SS data. The results showed that the integration of CT-SS data with demographics, risk factors, clinical manifestations, and laboratory parameters improved the prognostic performances of the ML models. Also, the kNN model had the best performance among the other three ML techniques with the accuracy of 94.1%, sensitivity of 100.0%, precision of 89.5%, specificity of 88.3%, and AUC around 97.2%. MLP and J48 decision tree models had a good prediction performance (AUCs > 93%), and their prognostic efficiency was better than the SVM model.

In several studies, ML techniques were evaluated in predicting mortality of the COVID-19 patients. In Moulaei et al. study [14], J48 decision tree, MLP, kNN, random forest (RF), and SVM data mining models were evaluated to predict the mortality of Covid-19 patients. Sixteen factors including demographics, risk factors, and clinical manifestations were used to predict the mortality of COVID-19 patients. The retrospective analysis of the data of 850 COVID-19 hospitalized patients showed that all ML algorithms have an acceptable prognostic performance (AUCs > 96%). RF model showed a slightly better performance and SVM was the weakest method in predicting mortality. In a similar paper conducted by Moulaei K et al. [15], mortality prediction of seven ML algorithms including the J48 decision tree, RF, kNN, MLP, Naïve Bayes (NB), eXtreme gradient boosting (XGBoost), and logistic regression (LR) models were evaluated. The results of this study had also shown that the RF model with the accuracy of 95.03%, sensitivity of 90.70%, precision of 94.23%, specificity of 95.10%, and AUC of 99.02 had better performance than other ML algorithms. In both studies, after the random forest model, kNN, MLP, and J48 decision tree algorithms had, respectively, the best prognostic performances. Their results were in close agreement with our findings.

In another study, the performances of LR, RF, SVM, and XGBoost algorithms for mortality prediction of COVID-19 patients were evaluated by Yadaw et al. [13]. In this study, data of 3841 confirmed COVID-19 patients (demographics, risk factors, and clinical manifestations) were analyzed and the results showed that the XGBoost algorithm with AUC of 91% was the best predictive model among all the models. In Gao et al. study [12], the retrospective analysis of 2520 COVID-19 hospitalized patients demonstrated that the model developed by the neural network (NN) had better performance in predicting COVID-19 patient’s physiological deterioration and death than LR, SVM, and gradient boosted decision tree(AUC = 97.60%).

These studies showed that the ML approach can help healthcare providers, clinicians, and health policymakers to timely predict the deterioration of the patient's condition and reduce the severe complications and the resulting mortalities. The integration of CT-SS data with demographics, risk factors, clinical manifestations, and laboratory parameters would improve the performances of the ML models for mortality prediction of COVID-19 patients and increase the survival rate of the patients. An ML model with this comprehensive collection of predictors could identify high-risk patients more efficiently and lead to the optimal use of hospital resources.

Limitations

This study had several limitations that must be addressed. (1) There were irregularities and imbalances in the registered data. Thus, noise and inadequate records were eliminated as much as possible. (2) As was mentioned, there were significantly higher numbers of records in the survived group than in the dead class (108 vs. 707). To solve this problem, SMOTE was used to minimize the bias via class balancing. (3) Unlike the prior studies in which there is a lack of radiological and imaging indicators; in this study, CT-SS data were used along with the demographics, risk factors, clinical manifestations, and laboratory results. Therefore, a more comprehensive collection of features was achieved that could enhance the prognostic performances of the models. (4) This is a retrospective single-center study that was conducted using the database collected at a designated referral hospital to deliver special healthcare services for COVID- 19 patients. Further studies need to be carried out with bigger and multicenter databases to perform the external validation of the proposed model. (5) In this study, clinical variables available at the initial time of admission were used to predict the mortality of COVID-19 patients. The time span from infection to admission that would affect the features was unclear. The evaluation of dynamic variations in the features could help for more efficient identification of patients with poor outcomes. 6) The prognostic performances of four well-known ML algorithms were evaluated. In the future, the performance accuracy of algorithms other than these algorithms would be compared to determine the best model to predict the mortality of COVID-19 patients.

Conclusions

In this study, we evaluated the prognostic significance of CT-SS in the mortality prediction of COVID-19 patients using four well-known ML algorithms including kNN, MLP, SVM, and J48 decision tree algorithms in the presence and absence of CT-SS data. The results showed that the performances of the algorithms were improved when they are fed by the dataset with CT-SS data and the kNN algorithm yielded better performance to predict the mortality of COVID-19 patients than other ML algorithms. An ML model with a comprehensive collection of predictors could identify high-risk patients more efficiently and lead to the optimal use of hospital resources. This optimal predictive approach can help healthcare providers, clinicians, and health policymakers to timely predict the deterioration of the patient's condition and reduce the severe complications and the resulting mortalities.

Availability of data and materials

All data generated and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request and Ilam University of Medical Sciences’ approval.

Abbreviations

COVID-19:

Coronavirus disease 2019

SARS-CoV-2:

Severe acute respiratory syndrome coronavirus 2

CT-SS:

Chest CT severity score

CT:

Computed Tomography

RT-PCR:

Reverse-transcriptase polymerase chain reaction

ROC:

Receiver operator characteristic curve

AUC:

Area under the ROC curve

ML:

Machine learning

k-NN:

K-nearest neighborhood

MLP:

Multilayer Perceptron

WEKA:

Waikato Environment for Knowledge Analysis

SMOTE:

Synthetic minority over-sampling technique.

OR:

Odds ratio

References

  1. Mohanty SK, Satapathy A, Naidu MM et al (2020) Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) and coronavirus disease 19 (COVID-19)—anatomic pathology perspective on current knowledge. Diagn Pathol 15(1):1–17

    Article  Google Scholar 

  2. Marco C, Michael R, Arturo C, Scott CD, Raffaela DN (2020) Features, evaluation and treatment coronavirus (COVID-19). StatPearls [Internet] StatPearls Publishing, Treasure Island

  3. Zakariaee SS, Abdi AI (2022) Influence of threshold selection strategy on the prognostic accuracy of chest CT severity score for mortality prediction of COVID-19 patients. Heart Lung 56:74–75

    Article  PubMed  PubMed Central  Google Scholar 

  4. Sohrabi C, Alsafi Z, O’neill N et al (2020) World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19). Int J Surg 76:71–76

    Article  PubMed Central  Google Scholar 

  5. Santura I, Kawalec P, Furman M, Bochenek T (2021) Chest computed tomography versus RT-PCR in early diagnostics of COVID-19—a systematic review with meta-analysis. Pol J Radiol 86(1):518–531

    Article  Google Scholar 

  6. Rokni M, Ghasemi V, Tavakoli Z (2020) Immune responses and pathogenesis of SARS-CoV-2 during an outbreak in Iran: comparison with SARS and MERS. Rev Med Virol 30(3):e2107

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Smith EM, Lee ACW, Smith JM, Thiele A, Zeleznik H, Ohtake PJ (2021) COVID-19 and post-intensive care syndrome: community-based care for ICU survivors. Home Health Care Manag Pract 33(2):117–124

    Article  Google Scholar 

  8. Alimohamadi Y, Tola HH, Abbasi-Ghahramanloo A, Janani M, Sepandi M (2021) Case fatality rate of COVID-19: a systematic review and meta-analysis. J Prev Med Hyg 62(2):E311

    PubMed  PubMed Central  Google Scholar 

  9. Wynants L, Van Calster B, Collins GS, et al (2020) Prediction models for diagnosis and prognosis of covid-19: systematic review and critical appraisal. BMJ 369.

  10. Wu G, Yang P, Xie Y et al (2020) Development of a clinical decision support system for severity risk prediction and triage of COVID-19 patients at hospital admission: an international multicentre study. Eur Respir J 56(2)

  11. Wong ZS, Zhou J, Zhang Q (2019) Artificial intelligence for infectious disease big data analytics. Infect Dis Health 24(1):44–48

    Article  PubMed  Google Scholar 

  12. Gao Y, Cai G-Y, Fang W et al (2020) Machine learning based early warning system enables accurate mortality risk prediction for COVID-19. Nat Commun 11(1):1–10

    Article  Google Scholar 

  13. Yadaw AS, Li Y-c, Bose S, Iyengar R, Bunyavanich S, Pandey G (2020) Clinical features of COVID-19 mortality: development and validation of a clinical prediction model. Lancet Digit Health 2(10):516–525

    Article  Google Scholar 

  14. Moulaei K, Ghasemian F, Bahaadinbeigy K, Sarbi RE, Taghiabad ZM (2021) Predicting mortality of COVID-19 patients based on data mining techniques. J Biomed Phys Eng 11(5):653

    Article  PubMed  PubMed Central  Google Scholar 

  15. Moulaei K, Shanbehzadeh M, Mohammadi-Taghiabad Z, Kazemi-Arpanahi H (2022) Comparing machine learning algorithms for predicting COVID-19 mortality. BMC Med Inform Decis Mak 22(1):1–12

    Article  Google Scholar 

  16. Zakariaee SS, Naderi N, Rezaee D (2022) Prognostic accuracy of visual lung damage computed tomography score for mortality prediction in patients with COVID-19 pneumonia: a systematic review and meta-analysis. Egypt J Radiol Nucl Med 53(1):1–9

    Article  Google Scholar 

  17. Zakariaee SS, Salmanipour H, Naderi N, Kazemi-Arpanahi H, Shanbehzadeh M (2020) Association of chest CT severity score with mortality of COVID-19 patients: a systematic review and meta-analysis. Clin Transl Imaging 1–14.

  18. García S, Luengo J, Herrera F (2015) Data preprocessing in data mining. Springer, New York

    Book  Google Scholar 

  19. Li J, Cheng K, Wang S et al (2017) Feature selection: a data perspective. ACM Comput Surv 50(6):1–45

    Article  Google Scholar 

  20. Saeys Y, Inza I, Larranaga P (2007) A review of feature selection techniques in bioinformatics. Bioinform 23(19):2507–2517

    Article  CAS  Google Scholar 

  21. Mei X, Lee H-C, Diao K-y et al (2020) Artificial intelligence–enabled rapid diagnosis of patients with COVID-19. Nat Med 26(8):1224–1228

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Jin C, Chen W, Cao Y et al (2020) Development and evaluation of an artificial intelligence system for COVID-19 diagnosis. Nat Commun 11(1):1–14

    Article  Google Scholar 

  23. Albahri AS, Hamid RA, Al-qays Z et al (2020) Role of biological data mining and machine learning techniques in detecting and diagnosing the novel coronavirus (COVID-19): a systematic review. J Med Syst 44(7):1–11

    Article  Google Scholar 

  24. Sadhana S, Pandiarajan S, Sivaraman E, Daniel D (2021) AI-based power screening solution for SARS-CoV2 infection: a sociodemographic survey and COVID-19 cough detector. Procedia Comput Sci 194:255–271

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Galván P, Fusillo J, González F et al (2021) Rapid screening for COVID-19 by applying artificial intelligence to chest computed tomography images: A feasibility study. Med Access Point Care 5:23992026211013644

    Article  PubMed  PubMed Central  Google Scholar 

  26. Liang H, Guo Y, Chen X et al (2022) Artificial intelligence for stepwise diagnosis and monitoring of COVID-19. Eur Radiol 32(4):2235–2245

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Gupta G, Salehi AW, Sharma B, Kumar N, Vaidya P (2022) COVID-19: automated detection and monitoring of patients worldwide using machine learning. In: Modeling, control and drug development for COVID-19 outbreak prevention. Springer, pp 731–61

  28. Zhao Z, Chen A, Hou W et al (2020) Prediction model and risk scores of ICU admission and mortality in COVID-19. PLoS ONE 15(7):e0236618

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Hu H, Yao N, Qiu Y (2020) Comparing rapid scoring systems in mortality prediction of critically ill patients with novel coronavirus disease. Acad Emerg Med 27(6):461–468

    Article  PubMed  PubMed Central  Google Scholar 

  30. Ryan L, Lam C, Mataraso S et al (2020) Mortality prediction model for the triage of COVID-19, pneumonia, and mechanically ventilated ICU patients: a retrospective study. Ann Med Surg 59:207–216

    Article  Google Scholar 

  31. Yan L, Zhang H-T, Goncalves J et al (2020) An interpretable mortality prediction model for COVID-19 patients. Nat Mach Intell 2(5):283–288

    Article  Google Scholar 

  32. Booth AL, Abels E, McCaffrey P (2021) Development of a prognostic model for mortality in COVID-19 infection using machine learning. Mod Pathol 34(3):522–531

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Das AK, Mishra S, Gopalan SS (2020) Predicting CoVID-19 community mortality risk using machine learning and development of an online prognostic tool. PeerJ 8:e10083

    Article  PubMed  PubMed Central  Google Scholar 

  34. Pan P, Li Y, Xiao Y et al (2020) Prognostic assessment of COVID-19 in the intensive care unit by machine learning methods: model development and validation. J Med Internet Res 22(11):e23128

    Article  PubMed  PubMed Central  Google Scholar 

  35. Allenbach Y, Saadoun D, Maalouf G et al (2020) Development of a multivariate prediction model of intensive care unit transfer or death: A French prospective cohort study of hospitalized COVID-19 patients. PLoS ONE 15(10):e0240711

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Assaf D, Ya G, Neuman Y et al (2020) Utilization of machine-learning models to accurately predict the risk for critical COVID-19. Intern Emerg Med 15(8):1435–1443

    Article  PubMed  PubMed Central  Google Scholar 

  37. Zhang Y, Xin Y, Li Q et al (2017) Empirical study of seven data mining algorithms on different characteristics of datasets for biomedical classification applications. Biomed Eng Online 16(1):1–15

    Article  PubMed  PubMed Central  Google Scholar 

  38. Zhou Y, He Y, Yang H et al (2020) Exploiting an early warning Nomogram for predicting the risk of ICU admission in patients with COVID-19: a multi-center study in China. Scand J Trauma Resusc Emerg Med 28(1):1–13

    Article  Google Scholar 

  39. Chin V, Samia NI, Marchant R et al (2020) A case study in model failure? COVID-19 daily deaths and ICU bed utilisation predictions in New York state. Eur J Epidemiol 35(8):733–742

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The authors would like to thank the Research Deputy of Ilam University of Medical Sciences and we also would like to thank all experts who participated in this study.

Funding

This research has been supported by Ilam University of Medical Sciences (approved number: IR.MEDILAM.REC.1401.255).

Author information

Authors and Affiliations

Authors

Contributions

SSZ contributed to conceptualization; Data gathering; Formal analysis; Investigation; Project administration; Funding acquisition; Methodology; Resources; Supervision; Writing, review, and editing—original draft. AIA contributed to conceptualization; Formal analysis; Investigation; Methodology; Writing, review, and editing—original draft. NN, MB contributed to data gathering; Investigation; Resources; Writing, review, and editing—original draft. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Seyed Salman Zakariaee.

Ethics declarations

Ethics approval and consent to participate

This article is extracted from a research project supported by Ilam University of Medical Sciences and all experimental protocols were approved by the ethical committee of Ilam University of Medical Sciences(approved number: IR.MEDILAM.REC.1401.255). All methods of the study were performed in accordance with the relevant guidelines and regulations of the ethical committee of Ilam University of Medical Sciences. Participation was voluntary and informed consent was obtained from all subjects and/or their legal guardians. Participants had the right to withdraw from the study at any time.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zakariaee, S.S., Abdi, A.I., Naderi, N. et al. Prognostic significance of chest CT severity score in mortality prediction of COVID-19 patients, a machine learning study. Egypt J Radiol Nucl Med 54, 73 (2023). https://doi.org/10.1186/s43055-023-01022-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s43055-023-01022-z

Keywords