Foetos: an expert system for fetal assessment

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine
• Obstetrics and Gynecology

Background:

• High-risk pregnancies require continuous fetal monitoring using various assessment tests.
• Interpreting complex fetal assessment data (fetal biophysical profile, contraction stress test, nonstress tests) can be challenging for clinicians.

Purpose of the Study:

• To describe Foetos, an expert system developed to assist clinical personnel in interpreting fetal assessment tests.
• To evaluate the effectiveness and agreement of Foetos' interpretations with clinical management.

Main Methods:

• Development of the Foetos expert system using the Genie knowledge engineering tool.
• Implementation of a mixed frame- and rule-based approach for clinical knowledge representation.
• Inclusion of diagnostic, prognostic, and therapeutic structures informed by heuristic interpretation and contextual clinical data.

Main Results:

• Foetos demonstrated a substantial level of agreement between its recommendations and actual clinical management.
• Initial retrospective and prospective validation studies support the system's utility.
• The system effectively integrates heuristic interpretations with the broader clinical context.

Conclusions:

• Expert systems like Foetos show promise in aiding clinical decision-making for fetal assessment.
• The obstetrical field presents a valuable area for the application of artificial intelligence (AI) techniques.
• Foetos' success suggests potential for AI to improve the interpretation of complex medical data in high-risk pregnancies.