Fig. 6

Comparing Bayesian Information Criterion score in predicting histopathology. The table presents the BIC score of each regression model in predicting histopathology. Lower score suggests better model performance. Color shade represents the magnitude of the evidence between the best and second-best models. Darker shade denotes stronger evidence supporting the model superiority. The column chart depicts the absolute value of the differences in BIC score among the three regression models: model 1—SNF clusters (four categories); model 2—disease conditions (six categories); and model 3—disease conditions, where we combined the two disease control groups and the FGR + HDP with the severe PE groups to make an equal number of groups (four categories). BIC, Bayesian Information Criterion; MVM, maternal vascular malperfusion; FVM, fetal vascular malperfusion; AI, acute inflammation; CI, chronic inflammation; DVH, distal villous hypoplasia; AVM, accelerated villous maturation