Angle of Progression (AoP) is a critical parameter for clinical assessment of fetal head descent and prediction of delivery mode, traditionally measured manually by experienced clinicians, which leads to efficiency and consistency issues. In this paper, we present a heatmap regression-based keypoint detection method as a baseline approach for the Intrapartum Ultrasound Grand Challenge (IUGC) 2025, designed to automatically measure the AoP in intrapartum ultrasound images. We employ a U-Net architecture for heatmap prediction to directly identify the three key points required for AoP measurement, followed by post-processing to extract precise coordinates. The method was evaluated on the IUGC 2025 dataset, trained with 300 annotated samples and tested on 501 samples, achieving an average AoP error of \(8.37^{\circ }\) and a MRE of 21.83 pixels. As a baseline method, we discuss current limitations and propose improvement directions, including semi-supervised learning to leverage unlabeled data, adoption of more advanced network architectures, and optimization of post-processing techniques. This study demonstrates the feasibility of automated AoP measurement in obstetric ultrasound imaging, potentially improving decision support tools in obstetric clinical practice.