The role of machine learning in infectious disease early detection and prediction in the MENA region: A systematic review

The role of machine learning in infectious disease early detection and prediction in the MENA region: A systematic review

This systematic review analyzes the implementation and effectiveness of machine learning (ML) approaches for infectious disease surveillance and prediction across the Middle East and North Africa (MENA) region. Adhering to PRISMA guidelines, studies published between 2016 and 2024 were examined to a...

Full description

Saved in:
Bibliographic Details
Main Authors: Radwan Qasrawi, Ghada Issa, Suliman Thwib, Razan AbuGhoush, Malak Amro, Raghad Ayyad, Stephanny Vicuna, Eman Badran, Yousef Khader, Raeda Al Qutob, Faris Al Bakri, Hana Trigui, Elie Sokhn, Emmanuel Musa, Jude Dzevela Kong
Format: Article
Language:English
Published: Elsevier 2025-01-01
Series:Informatics in Medicine Unlocked
Subjects:
Artificial intelligence (AI)
Machine learning (ML)
infectious diseases
Early detection
Outbreak prediction
Medical imaging
Online Access:http://www.sciencedirect.com/science/article/pii/S2352914825000395
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This systematic review analyzes the implementation and effectiveness of machine learning (ML) approaches for infectious disease surveillance and prediction across the Middle East and North Africa (MENA) region. Adhering to PRISMA guidelines, studies published between 2016 and 2024 were examined to assess model structures, performance metrics, and dataset characteristics. The findings reveal a predominance of deep learning approaches, particularly Convolutional Neural Networks (CNNs), achieving mean accuracy rates of 96.3 % in pathogen detection from medical imaging. Random Forest algorithms demonstrated superior performance in disease outbreak prediction, with mean ACC scores of 0.85. Iran, Saudi Arabia, and Egypt emerged as regional leaders, collectively contributing 54 % of the analyzed studies. The temporal analysis showed peak research output in 2022 (n = 30 studies), followed by a 25 % decline in 2023. Despite promising performance, challenges such as data quality, infrastructural limitations, and algorithmic bias persist. This review highlights the need for standardized protocols, enhanced digital infrastructure, and collaborative efforts to realize the full potential of ML in enhancing public health interventions across the region. Future research directions should prioritize multi-center validation studies, standardized reporting frameworks, and integration of diverse data modalities to enhance model robustness and clinical applicability.
ISSN:2352-9148

Similar Items