Artificial intelligence and machine learning in cardiotocography: A scoping review.

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Fetal Monitoring

Background:

• Artificial intelligence (AI) offers potential to enhance medical data analysis by identifying patterns.
• AI in cardiotocography (CTG) interpretation aims to reduce observer variability and bias.
• Current CTG interpretation faces challenges with inter- and intra-observer variability.

Purpose of the Study:

• To systematically review and map the current landscape of AI-assisted CTG interpretation.
• To identify and present various AI approaches, detailing their advantages and limitations.
• To provide an overview of the current state of AI in CTG analysis.

Main Methods:

• A comprehensive literature search was conducted on Ovid Medline and PubMed.
• The systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Scoping Reviews (PRISMA-ScR) guidelines.
• Studies investigating AI algorithms or systems for CTG classification were included.

Main Results:

• Forty studies evaluating AI algorithms for CTG classification were summarized.
• The Oxford Sonicaid system, a widely used clinical tool, was also included.
• A variety of AI approaches for CTG interpretation were identified.

Conclusions:

• Promising AI-based approaches for CTG interpretation exist but lack widespread clinical acceptance.
• Further research and algorithm refinement are necessary to develop validated decision-support systems.
• The need for larger, curated, and labeled datasets is highlighted for advancing AI in CTG.