HUPatho2BrainTumor

HUPatho2BrainTumor

This dataset contains histopathological images of brain tumors for use in machine learning classification models and research purposes.

Dataset Overview

The dataset is organized into 6 folders representing different brain tumor types and background brain tissue:

Folder
Tumor Type
Cases
Images

G
Glioblastoma, IDH-wild type
66
1,146

A
Astrocytoma, IDH-mutant
26
248

O
Oligodendroglioma, IDH-mutant, 1p19q-codeleted.
24
272

M
Brain Metastases
81
1,041

L
Primary Central Neural System Lymphoma
52
491

B
Background Brain
147(augmented from 5 autopsy cases)
4,828

File Naming Convention

Each folder contains images named as: `[Category][CaseNumber]_[ImageNumber].jpg`

Example: `G0001_01.jpg` represents the first image of the first glioblastoma case

Usage

This dataset is suitable for:

Training and evaluating brain tumor classification models

Deep learning research in digital pathology

Computer-aided diagnosis system development

Medical image analysis studies

Image Specifications

Format: JPEG

Multiple images per case to capture different regions and features

Notes

Case identifiers have been anonymized for privacy protection

The dataset includes various augmentations for the background brain category to enhance model training

Feel free to use this dataset for developing classification models and advancing research in computational pathology!