# PAD-UFES-20: smartphone photographs of 1,641 skin lesions from 1,373 patients with clinical metadata (UFES, Brazil)

## Facts

- Provider: Andre G. C. Pacheco, Renato A. Krohling et al. (UFES); repackaged by Chehab Lab
- License: CC-BY-4.0 (https://data.mendeley.com/public-api/datasets/zr7vgbcyr2?version=1)
- Size: 10 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/chehablab/PAD-UFES-20
- Access: public, readable without an account
- DOI: https://doi.org/10.17632/zr7vgbcyr2.1
- Category: Health & Biomedical
- Source: https://doi.org/10.1016/j.dib.2020.106221
- Page: https://scigantic.com/archive/5182db11-ae25-5942-9a53-a439c23a1e44
- Keywords: Health & Biomedical, Parquet, Hugging Face

## About

PAD-UFES-20 is a skin cancer dataset of clinical photographs taken with smartphones by the Dermatological and Surgical Assistance Program (PAD) of the Federal University of Espirito Santo (UFES), Brazil, rather than with a dermatoscope. It has 2,298 images of 1,641 lesions from 1,373 patients across six diagnoses: actinic keratosis (ACK), basal cell carcinoma (BCC), melanoma (MEL), nevus (NEV), squamous cell carcinoma (SCC) and seborrheic keratosis (SEK). PAD is a nonprofit program that treats skin lesions, mainly for low-income patients. Each image comes with the patient and lesion fields of the original metadata.csv, such as age, lesion location, Fitzpatrick skin type, lesion diameters and whether the diagnosis was biopsy-proven. The repackaging keeps the original PNGs unchanged and adds an integer label; split by patient_id to avoid leakage. Eight Parquet files under data/, about 3.6 GB (measured 2026-11-21); read with Hugging Face datasets. Use for multimodal skin lesion classification combining images and patient data. Original data on Mendeley Data (DOI 10.17632/zr7vgbcyr2.1), released under CC BY 4.0. Cite Pacheco et al., Data in Brief 32:106221 (2020).

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("chehablab/PAD-UFES-20")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `data/train-00000-of-00008.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/5182db11-ae25-5942-9a53-a439c23a1e44
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/5182db11-ae25-5942-9a53-a439c23a1e44
- Over MCP: `get_schema_card` with id `5182db11-ae25-5942-9a53-a439c23a1e44` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Health & Biomedical datasets](https://scigantic.com/topics/health-and-biomedical)
