Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test
Background: Needless to say that correct and real-time detection and effective prognosis of the COVID-19 are necessary measures to deliver the best possible care for patients and, accordingly, diminish the pressure on the health care industries. The main purpose of the present paper was to devise p...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tehran University of Medical Sciences
2022-03-01
|
| Series: | Journal of Biostatistics and Epidemiology |
| Subjects: | |
| Online Access: | https://jbe.tums.ac.ir/index.php/jbe/article/view/704 |
| _version_ | 1811280323608051712 |
|---|---|
| author | Mohammad Reza Afrash Leila Erfanniya Morteza Amraei Nahid Mehrabi Saeed Jelvay raoof Nopour Mostafa Shanbehzadeh |
| author_facet | Mohammad Reza Afrash Leila Erfanniya Morteza Amraei Nahid Mehrabi Saeed Jelvay raoof Nopour Mostafa Shanbehzadeh |
| author_sort | Mohammad Reza Afrash |
| collection | DOAJ |
| description |
Background: Needless to say that correct and real-time detection and effective prognosis of the COVID-19 are necessary measures to deliver the best possible care for patients and, accordingly, diminish the pressure on the health care industries. The main purpose of the present paper was to devise practical solutions based on Machine Learning (ML) techniques to ease the COVID-19 screening in routine blood test data. We came up with different algorithms for the early detection of COVID-19 and finally succeeded to opt for the best performing algorithm.
Material and methods: In this developmental study, the laboratory data of 1703 COVID-19 and non-COVID-19 patients Using a single-center registry from February 9, 2020, to December 20, 2020, were analyzed by several ML algorithms which included, K-Nearest Neighbors, AdaBoost Classifier, Decision Tree, and HistGradient Boosting Classifier. A 10-fold cross-validation method and six performance evaluation metrics were used to evaluate and compare these implemented algorithms. Using the best ML-developed model, a Clinical Decision Support System (CDSS) was implemented with C# programming language.
Results: The results indicated that the best performance belongs to the AdaBoost classifier with mean accuracy, specificity, sensitivity, F-measure, KAPA rate, and ROC of 87.1%, 85.3%, 87.3%, 87.1 %, 89.4%, and 87.3 % respectively
Discussion: The ML makes a reasonable level of accuracy possible for an early diagnosis and screening of COVID-19. The empirical results reveal that the Adaboost model yielded higher performance compared with other classification models and was used for developing our CDSS interface in discriminates positive COVID-19 from negative cases.
|
| first_indexed | 2024-04-13T01:11:33Z |
| format | Article |
| id | doaj.art-e103027b5582448d9c4a47e057dbdd1b |
| institution | Directory Open Access Journal |
| issn | 2383-4196 2383-420X |
| language | English |
| last_indexed | 2024-04-13T01:11:33Z |
| publishDate | 2022-03-01 |
| publisher | Tehran University of Medical Sciences |
| record_format | Article |
| series | Journal of Biostatistics and Epidemiology |
| spelling | doaj.art-e103027b5582448d9c4a47e057dbdd1b2022-12-22T03:09:08ZengTehran University of Medical SciencesJournal of Biostatistics and Epidemiology2383-41962383-420X2022-03-0181Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood testMohammad Reza Afrash0Leila Erfanniya1Morteza Amraei2Nahid Mehrabi3Saeed Jelvay4raoof Nopour5Mostafa Shanbehzadeh6Student Research Committee, School of Allied Medical Sciences, Shahid Beheshti University of Medical Sciences, Tehran, IranAssistant Professor of Health Information Management, Department of Health Information Technology, Faculty of Paramedical, Zaheda University of Medical Sciences, Zahedan, IranAssistant Professor of Health Information Management, Clinical Education Research Center, Shiraz University of Medical Sciences, Shiraz, IranAssistant Professor, Department of Health Information Technology, Aja University of MedicaInstructor of Health Information Technology, Abadan University of Medical Sciences, Abadan, Iran2. MSc of Health Information Technology, Department of Health Information Technology and Management, School of Allied Medical Sciences, Tehran University of Medical Sciences, Tehran, Iran- Department of Health Information Technology, School of Paramedical, Ilam University of Medical Sciences, Ilam, Iran Background: Needless to say that correct and real-time detection and effective prognosis of the COVID-19 are necessary measures to deliver the best possible care for patients and, accordingly, diminish the pressure on the health care industries. The main purpose of the present paper was to devise practical solutions based on Machine Learning (ML) techniques to ease the COVID-19 screening in routine blood test data. We came up with different algorithms for the early detection of COVID-19 and finally succeeded to opt for the best performing algorithm. Material and methods: In this developmental study, the laboratory data of 1703 COVID-19 and non-COVID-19 patients Using a single-center registry from February 9, 2020, to December 20, 2020, were analyzed by several ML algorithms which included, K-Nearest Neighbors, AdaBoost Classifier, Decision Tree, and HistGradient Boosting Classifier. A 10-fold cross-validation method and six performance evaluation metrics were used to evaluate and compare these implemented algorithms. Using the best ML-developed model, a Clinical Decision Support System (CDSS) was implemented with C# programming language. Results: The results indicated that the best performance belongs to the AdaBoost classifier with mean accuracy, specificity, sensitivity, F-measure, KAPA rate, and ROC of 87.1%, 85.3%, 87.3%, 87.1 %, 89.4%, and 87.3 % respectively Discussion: The ML makes a reasonable level of accuracy possible for an early diagnosis and screening of COVID-19. The empirical results reveal that the Adaboost model yielded higher performance compared with other classification models and was used for developing our CDSS interface in discriminates positive COVID-19 from negative cases. https://jbe.tums.ac.ir/index.php/jbe/article/view/704: COVID-19CoronavirusMachine learningArtificial intelligenceDecision Support Systems |
| spellingShingle | Mohammad Reza Afrash Leila Erfanniya Morteza Amraei Nahid Mehrabi Saeed Jelvay raoof Nopour Mostafa Shanbehzadeh Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test Journal of Biostatistics and Epidemiology : COVID-19 Coronavirus Machine learning Artificial intelligence Decision Support Systems |
| title | Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test |
| title_full | Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test |
| title_fullStr | Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test |
| title_full_unstemmed | Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test |
| title_short | Machine Learning-Based Clinical Decision Support System for automatic diagnosis of COVID-19 based on the routine blood test |
| title_sort | machine learning based clinical decision support system for automatic diagnosis of covid 19 based on the routine blood test |
| topic | : COVID-19 Coronavirus Machine learning Artificial intelligence Decision Support Systems |
| url | https://jbe.tums.ac.ir/index.php/jbe/article/view/704 |
| work_keys_str_mv | AT mohammadrezaafrash machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT leilaerfanniya machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT mortezaamraei machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT nahidmehrabi machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT saeedjelvay machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT raoofnopour machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest AT mostafashanbehzadeh machinelearningbasedclinicaldecisionsupportsystemforautomaticdiagnosisofcovid19basedontheroutinebloodtest |