Artificial Intelligence in Pediatric Diagnostics: A Systematic Review of Accuracy, Safety, and Clinical Impact

Isra Mhamedahmed¹, Afag  Badr², Hiba  Hajali³, Bensiradj Mounir Abdelhak⁴, Abdulwahhab Al-Shaikhli⁵*

Authors

Keywords:

Artificial intelligence, Paediatrics, Diagnostic accuracy, Machine learning, Clinical safety, Systematic review

Abstract

DOI : 10.5281/zenodo.16934731


Corrsponding Author : Afag  Badr

Background Artificial intelligence (AI) is increasingly being adopted in paediatric diagnostics, offering potential benefits in diagnostic speed and accuracy. However, its clinical safety, validation, and applicability to diverse paediatric populations remain underexplored.

 

Objective:

This systematic review aimed to evaluate the diagnostic accuracy, clinical safety, and implementation challenges of AI tools used in paediatric diagnostics

 

Methods:

A comprehensive literature search was conducted across PubMed, Scopus, IEEE Xplore, and Web of Science for studies published between 2005 and 2025. Eligible studies evaluated AI-based diagnostic tools in paediatric populations (0–18 years) and reported performance metrics such as sensitivity, specificity, and area under the curve (AUC). Quality was assessed using the QUADAS-2 tool, and a narrative synthesis was performed due to methodological heterogeneity.

Results:

Forty-two studies were included, covering a wide range of AI algorithms and paediatric conditions including respiratory disorders, neurological conditions, ophthalmological diseases, dermatology, oncology, and cardiology. Most AI models demonstrated high diagnostic performance, with AUC values commonly exceeding 0.90. However, the majority of studies lacked external validation, were single-centre, and did not report clinical outcomes or adverse events. Ethical concerns, including data bias and lack of explainability, were noted but infrequently addressed empirically.

Conclusion:

AI-based diagnostic tools show strong promise in enhancing paediatric diagnostics, particularly for image-based conditions. However, significant gaps remain in safety reporting, real-world validation, and ethical oversight. Rigorous prospective trials and clinician-AI integration strategies are essential for their responsible deployment in paediatric care.



Downloads

Download data is not yet available.
oooo

Downloads

Published

2025-08-26

Most read articles by the same author(s)

Similar Articles

1-10 of 44

You may also start an advanced similarity search for this article.