Russian Scientists Develop AI Model to Predict Preterm Birth Risks
The CatBoost-based system analyzes electronic health records to support early intervention and personalized care.
Researchers at Petrozavodsk State University (PetrSU) have developed an artificial intelligence model based on CatBoost to assess the risk of preterm birth at an early stage. The system uses data from electronic medical records, Medvestnik reported.
Personalized Approach
Among 14 algorithms tested, the CatBoost model achieved 81% accuracy and 87% sensitivity, with text extraction completeness reaching 99.8%. The model was trained on 10,000 anonymized medical records using 54 parameters, including clinical and laboratory data.
The system analyzes factors such as placental insufficiency, infections, cervical insufficiency, multiple pregnancies and IVF. This enables physicians to identify risks earlier and initiate preventive measures.
The researchers note that preterm birth results from a complex interaction of multiple factors unique to each patient. This makes a personalized approach essential, based on integrating a wide range of clinical data.
