An Automated Machine Learning Classification Model for Predicting Placental Abruption

Placental abruption is detachment of the placenta before delivery from the implantation site that may have a potential to develop life-threating emergency clinic syptoms. The multifactorial nature of this disorder and no lab testing or procedures that can diagnose placental abruption. makes it difficult to predict. Artificial intelligence (AI) and machine learning (ML) have the potential to enhance clinical decision-making and enable precise assessments. This study purposed on predictive 15 ML models for placental abruption high-lighting input characteristics, performance metrics, and validation. The medical records of 564 patients were analyzed between 2021 and 2025 for studies using AI to develop predictive models for placental abruption. Findings were analyzed with Python software and Pycaret library. The model integrated data for 5 variables (features) for the prediction. Among 15 machine learning algorithms, Logistic regression was chosen as the best model. The performance metrics were determined as follows: accuracy rate of 0.85, AUC of 0.91, recall of 0.85, precision of 0.85, and F1 score of 0.85. In the ranking based on their importance in the classification model, gestational age at delivery was observed to have the highest importance for classification. Twenty-eight unseen cases were utilized for an extra validation step. The model achieved a high accuracy on this set, with 21 cases correctly predicted. The presented 15 ML models in our study had significant accuracy in predicting placental abruption , but these models require further development before they can be applied in a clinical setting.