Between 6% and 40% of vaginal births may result in trauma to the pelvic floor muscles (PFM), leading to long-term consequences such as incontinence or prolapse [1]. Computational simulations can be used to biomechanically analyze the birthing process; however, the integration of these simulations into clinical practice is hindered by high computational costs. To overcome this limitation, surrogate models can replace the simulations, enabling faster results. This study aims to develop an AI framework for predicting PFM stretch during vaginal delivery. A finite element model of the PFM and fetal head was used to simulate childbirth. Material properties and muscle geometry were varied to achieve patient-specific characteristics. A dataset was created with 5618 completed simulations, with each pelvic floor node corresponding to an observation. Specifically, 49 nodes near the urogenital hiatus of the PFM were selected. The study utilized initial node coordinates, material properties, muscle diameters, node position, urogenital hiatus length, and time frame as training features. Decision Trees (DT), Random Forest (RF), Extreme Gradient Boosting (XGBT), and Artificial Neural Networks (ANN) were the selected machine learning models. The stretch was predicted for multiple instants of fetal descent to determine both its peak and the instant when it occurs. Preliminary results included a MAE of 0.239 mm for the DT model, 0.200 mm for the RF model, 0.544 mm for the XGBT model, and 0.397 mm for the ANN. The development of this framework consists of a preliminary step toward the implementation of patient-specific, near real-time computational simulations in a clinical environment. The ability to predict the stretch suffered by the woman on the pelvic floor immediately before or during childbirth could aid in medical decision-making and in the identification of non-visible injuries. [1] Kreft, M., Cai, P., Furrer, E., Richter, A., Zimmermann, R., and Kimmich, N. The evolution of levator ani muscle trauma over the first 9 months after vaginal birth. International Urogynecology Journal, 33(9):2445–2453, 2022. doi: 10.1007/s00192- 021-05034-z.