PSFHS: intrapartum transperineal ultrasound with pubic symphysis and fetal head segmentation, 1,358 images (MICCAI 2023)

Provider
Lu et al. (PSFHS authors) and Bai et al. (PS-FH-AOP challenge); repackaged by MedOtter
License
CC-BY-4.0 (source)
Size
3 files
Format
Parquet (in sampled files)
Stored at
Hugging Face (https://huggingface.co/datasets/MedOtter/ps-fh-aop-2023)
Access
Free and public. No account with the data provider is needed.
DOI
10.5281/zenodo.10969427
Category
Health & Biomedical
Source
https://doi.org/10.1038/s41597-024-03266-4

About this dataset

PSFHS (Pubic Symphysis-Fetal Head Segmentation) is an intrapartum ultrasound dataset for measuring how far the baby's head has descended during labor. Transperineal ultrasound in the mid-sagittal plane shows the mother's pubic symphysis and the fetal head; outlining both lets software compute the angle of progression (AoP), which helps predict whether vaginal delivery will succeed. This repository holds the 1,358 pixel-annotated images released on Zenodo for the MICCAI 2023 PS-FH-AOP grand challenge, each a 256x256 RGB ultrasound frame with a 3-class mask (0 background, 1 pubic symphysis, 2 fetal head), sample IDs 03744 to 05101, single train split. The challenge also used JNU-IFM images, which are not included. One Parquet file under data/, about 70 MB (measured 2026-11-21); read with Hugging Face datasets. Use for obstetric ultrasound segmentation and automated angle of progression measurement. Released on Zenodo (DOI 10.5281/zenodo.10969427) under CC BY 4.0. Cite Lu et al., Scientific Data 11:436 (2024), doi 10.1038/s41597-024-03266-4.

How to load this dataset in Python

from datasets import load_dataset
ds = load_dataset("MedOtter/ps-fh-aop-2023")
print(ds)

Data structure

File types in sampled files: Parquet

Files

Example file: data/train-00000-of-00001.parquet

Analyze this dataset without downloading it

On Scigantic, PSFHS: intrapartum transperineal ultrasound with pubic symphysis and fetal head segmentation, 1,358 images (MICCAI 2023) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Health & Biomedical datasets