Rapid Triage of Children with Suspected COVID-19 Using Laboratory-Based Machine-Learning Algorithms

Abstract

In order to limit the spread of the novel betacoronavirus (SARS-CoV-2), it is necessary to detect positive cases as soon as possible and isolate them. For this purpose, machine-learning algorithms, as a field of artificial intelligence, have been recognized as a promising tool. The aim of this study was to assess the utility of the most common machine-learning algorithms in the rapid triage of children with suspected COVID-19 using easily accessible and inexpensive laboratory parameters. A cross-sectional study was conducted on 566 children treated for respiratory diseases: 280 children with PCR-confirmed SARS-CoV-2 infection and 286 children with respiratory symptoms who were SARS-CoV-2 PCR-negative (control group). Six machine-learning algorithms, based on the blood laboratory data, were tested: random forest, support vector machine, linear discriminant analysis, artificial neural network, k-nearest neighbors, and decision tree. The training set was validated through stratified cross-validation, while the performance of each algorithm was confirmed by an independent test set. Random forest and support vector machine models demonstrated the highest accuracy of 85% and 82.1%, respectively. The models demonstrated better sensitivity than specificity and better negative predictive value than positive predictive value. The F1 score was higher for the random forest than for the support vector machine model, 85.2% and 82.3%, respectively. This study might have significant clinical applications, helping healthcare providers identify children with COVID-19 in the early stage, prior to PCR and/or antigen testing. Additionally, machine-learning algorithms could improve overall testing efficiency with no extra costs for the healthcare facility.

Diagnostic value of routine blood tests in differentiating between SARS-CoV-2, influenza A, and RSV infections in hospitalized children: a retrospective study

“…AI’s ability to analyze data, recognize patterns, and process datasets may bolster the accuracy of risk evaluations. Amid the COVID-19 crisis, successful initiatives have employed automated machine learning to distinguish between influenza virus infections and SARS-CoV-2 [ 27 ], as well as promptly identifying COVID-19 in children [ 28 , 29 ]. Advances have also been made in utilizing decision tree models based on hemogram outcomes to distinguish between RSV and COVID-19 in infants [ 30 ].…”

COVID-19 Pediatric Follow-Up: Respiratory Long COVID-Associated Comorbidities and Lung Ultrasound Alterations in a Cohort of Italian Children

“…The evolving landscape of medical research and technology opens up potential applications of artificial intelligence (AI) in the assessment and monitoring of pediatric patients with long COVID. Integrating AI into these processes holds the promise of enhancing diagnostic accuracy and offering valuable insights into the long-term respiratory effects [73][74][75].…”

Analysis of the Forces Driving Public Hospitals’ Operating Costs Using LMDI Decomposition: The Case of Japan

“…Second, information technologies such as 5G, big data analytic engines, remote medical consultation, and applications of Artificial Intelligence (AI) are expected to suppress the operating costs [70][71][72][73][74][75]. These measures help in downsizing the scale of public hospitals and improve the cost effectiveness [76,77]. Furthermore, policies such as cost-sharing can be proliferated with an aim to divert some of the inpatients from long-term stays in hospitals to community-based care facilities for conditions that do not necessitate medical intervention, especially in cases of elderly and chronic patients [78].…”