README.md

---
configs:
- config_name: default
  data_files:
  - split: train
    path:
      - "Training/1-notumor/*.jpg"
      - "Training/2-glioma/*.jpg"
      - "Training/3-meningioma/*.jpg"
      - "Training/4-pituitary/*.jpg"

  - split: test
    path:
      - "Testing/1-notumor/*.jpg"
      - "Testing/2-glioma/*.jpg"
      - "Testing/3-meningioma/*.jpg"
      - "Testing/4-pituitary/*.jpg"

task_categories:
- image-classification
language:
- en
size_categories:
- 1K<n<10K
---

## Dataset description

This dataset is a combination of the following three datasets :

[Figshare](https://figshare.com/articles/dataset/brain_tumor_dataset/1512427)\
[SARTAJ dataset](https://github.com/sartajbhuvaji/brain-tumor-classification-dataset)\
[Br35H](https://www.kaggle.com/datasets/ahmedhamada0/brain-tumor-detection?select=no)

This dataset contains 7023 images of human brain MRI images which are divided into 4 classes: glioma - meningioma - no tumor and pituitary.
No tumor class images were taken from the Br35H dataset.

## Acknowledgement

This dataset is reproduced and taken from [Kaggle](https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset)