Pubic Symphysis-Fetal Head Segmentation and Angle of Progression

During the process of labor, the intrapartum transperineal ultrasound examination serves as a valuable tool, allowing direct observation of the relative positional relationship between the pubic symphysis and fetal head (PSFH). Accurate assessment of fetal head descent and the prediction of the most suitable mode of delivery heavily rely on this relationship. However, achieving an objective and quantitative interpretation of the ultrasound images necessitates precise PSFH segmentation (PSFHS), a task that is both time-consuming and demanding. Integrating the potential of artificial intelligence (AI) in the field of medical ultrasound image segmentation, the development and evaluation of AI-based models rely significantly on access to comprehensive and meticulously annotated datasets. Unfortunately, publicly accessible datasets tailored for PSFHS are notably scarce. Bridging this critical gap, we introduce a PSFHS dataset comprising 1358 images, meticulously annotated at the pixel level. The annotation process adhered to standardized protocols and involved collaboration among medical experts. Remarkably, this dataset stands as the most expansive and comprehensive resource for PSFHS to date.

        All data records are available as files on the web page https://doi.org/10.5281/zenodo. 10901776. The unzipped file folder of this dataset contains the original transperineal ultrasound images and annotation ground truth images. The unzipped file is organized into 2 folders, named “image_mha” and “label_mha”, that contain original transperineal ultrasound images and corresponding ground truth images, respectively. The images in these 2 folders are stored, named and arranged according to the same rule, where a specific image in the “label_mha” folder is the ground truth of the image with the same name in the “image_mha” folder. Images are named as “n.mha”, where “n” means the number of images. In the dataset, there are 1358 images (”n” from 03744 to 05101) of 1124 pregnant women. The images in the “image_mha” folder contain pixels labelled as 0, 1, or 2, where 0 represents the background, 1 represents the PS, and 2 represents the FH. These data can be accessed using the software “Insight Segmentation and Registration Toolkit”, available at https://itk.org/.

         The whole dataset used for the PSFHS challenge of MICCAI2023 (https://ps-fh-aop-2023.grand-challenge.org/) includes two parts: one is this PSFHS dataset (https://doi.org/10.5281/zenodo.10901776) and another is from the JNU-IFM dataset (https://doi.org/10.6084/m9.figshare.14371652). These images in the PSFHS dataset can also be used for the Intrapartum Ultrasound Grand Challenge (IUGC) 2024 of MICCAI 2024 (https://codalab.lisn.upsaclay.fr/competitions/18413).