Artificial Intelligence in Paediatric Emergencies: A Narrative Review

Abstract

Background: The functionality of Artificial intelligence (AI) in paediatric practices has been gaining more attention for last five years. Since then, researchers have started observing that the techniques are helpful in dealing multiple facets of childhood diseases including emergency like situations. This article has been aimed to discuss the current status of usefulness of AI in paediatric emergencies. Methods: Total 22 research articles have been reviewed. Articles were searched from electronic database like Pubmed, Medline, Google scholar. Artificial intelligence, machine learning, paediatric emergencies, childhood diseases were the key words used to ease the search. Results: Out of 22, 15 were chosen as relatable to paediatric emergency situations per se. After reviewing the available literature, the utility of AI in paediatric emergencies had been discussed under four sub headings: i) Diagnosis; ii) Predictive modelling iii) Assistance in Antimicrobial stewardship iv) Management of emergency department resources. Conclusion: AI and different machine learning techniques have been proven as reliable accompaniment of paediatricians. They can provide their support in terms of early diagnosis for example the septic shock in children, prediction of disease severity like in the cases of traumatic brain injury, drug doses and emergency resource management. Lack of research on extensive data on far reaching population, legal and trust issues and unfriendly software’s are the challenges those need to be resolved for utilizing AI at its higher potential in paediatric healthcare.