Soil-transmitted Helminths and Schistosoma mansoni Eggs Image Dataset

Context

This dataset was developed as part of the study “Deep learning-based automated detection and multiclass classification of soil-transmitted helminths and Schistosoma mansoni eggs in fecal smear images” (Oyibo et al., 2025, Scientific Reports).

The work aimed to build a deep learning-ready image dataset to support the automated detection and classification of Soil-transmitted Helminths and Schistosoma mansoni parasite eggs using low-cost digital microscopy. Images were captured using the Schistoscope, a cost-effective automated microscope, and supplemented with the AI4NTD KK2.0 P3.0 STH & SCHm (https://doi.org/10.34740/kaggle/ds/3675429).

The combined dataset addresses limitations of previous collections by improving class balance, incorporating diverse field samples, and supporting AI model development for diagnostic use in low-resource settings.

Data Summary

Source locations: Schistoscope dataset captured during field studies conducted in the Federal Capital Territory, Nigeria.

Sample type: Fecal smears prepared using the Kato-Katz technique (41.7 mg template).

Image acquisition: Schistoscope configured with a 4× objective lens (0.10 NA).

Total slides imaged: 300 stool samples (Schistoscope dataset) and 272 slides (AI4NTD dataset).

Resolution: 2028 × 1520 pixels - RGB - (Schistoscope dataset)  and  1640 x 1232 pixels - RGB - (AI4NTD dataset).

Total field-of-view (FOV) images in combined dataset : 10,562.

Egg classes included:

Ascaris lumbricoides - 8,600 eggs

Trichuris trichiura - 4,083 eggs

Hookworm - 4,512 eggs

Schistosoma mansoni -3,920 eggs

Annotations: Bounding box labels provided in COCO formats.

File formats: PNG for images and JSON for annotations.

Intended use: Training and evaluation of machine learning and deep learning models for automated parasite egg detection and classification.

Ethics Statement

Ethical approval for the Schistoscope dataset collection was granted by the Federal Capital Territory (FCT) Health Research Ethics Committee, Nigeria, under approval number FHREC/2022/01/102/05-07-22.

All procedures were carried out in accordance with relevant guidelines and regulations.

Informed consent was obtained from the parents or guardians of participating school-age children.

No personally identifiable information is contained within the dataset.

Acknowledgements

We acknowledge the Neglected Tropical Disease (NTD) team from the Federal Ministry of Health and the FCT Public Health Department, Abuja, Nigeria, for facilitating the fieldwork.

We also thank the Delft University of Technology Global Initiative, Leiden University Medical Center, and the University of Lagos for technical support.

This work was funded by the NTD Innovation Prize and the NWO-WOTRO Science for Global Development Program (Grant Number W 07.30318.009, INSPiRED Project).

Citation

If you use this dataset, please cite both the dataset and the associated publication as follows:

Dataset citation:Oyibo, P., Meulah, B., Agbana, T., van Lieshout, L., Oyibo, W., Vdovin, G., & Diehl, J.-C. (2025). Soil-transmitted Helminths and Schistosoma mansoni Eggs Image Dataset [Data set]. Zenodo. https://doi.org/10.5281/zenodo.13843815

Associated publication:Oyibo, P., Meulah, B., Agbana, T., van Lieshout, L., Oyibo, W., Vdovin, G., & Diehl, J.-C. (2025). Deep learning-based automated detection and multiclass classification of soil-transmitted helminths and Schistosoma mansoni eggs in fecal smear images. Scientific Reports, 15(21495). https://doi.org/10.1038/s41598-025-02755-9