Datasets:
| license: cc-by-nc-sa-4.0 | |
| task_categories: | |
| - image-classification | |
| tags: | |
| - medical | |
| - dermatology | |
| size_categories: | |
| - 1K<n<10K | |
| # DermaConIN SkinDisNet ArsenicSkinImageBD | |
| ## Dataset Description | |
| This is a hybrid dermatology image classification dataset designed for identifying common skin conditions, including arsenic-related skin lesions. | |
| ### Data Composition | |
| The dataset is a combination of data from multiple sources: | |
| - **DermaConIN & SkinDisNet:** Includes skin diseases that have more than 100 images available in these respective datasets. | |
| - **Arsenic Skin Images:** | |
| - All positive arsenic skin lesion images are categorized into the `Arsenic` folder. | |
| - **Others Class:** | |
| - Contains negative images from the arsenic data. | |
| - Contains healthy skin images. | |
| - Contains all other skin diseases that had fewer than 100 images available in the DermaConIN and SkinDisNet datasets. | |
| ## Usage | |
| This dataset is structured for easy loading using the Hugging Face `datasets` library. The images are stored in optimized Parquet format for fast streaming and downloading. | |
| ```python | |
| from datasets import load_dataset | |
| dataset = load_dataset("actorcritic/DermaConIN_SkinDisNet_ArsenicSkinImageBD") | |
| ``` | |