Obstetrical decision-making based on predictive expert analysis

The validation results for the prognostic module of the obstetrical expert system NST-EXPERT are presented. The validation method consists of calculating the percentages of man/machine agreement for the expert system and a team of three perinatologists. This comparison was carried out for the following three categories: man neonatal outcome prediction vs. machine neonatal outcome prediction, man/machine outcome prediction vs. real outcome, and man/machine bad outcome prediction vs. real bad outcome. The results suggest that the use of prognostic models, such as that of NST-EXPERT, in obstetrical expert systems can improve the quality of clinical assistance in high-risk pregnancies.