BreakHis - Breast Cancer Histopathological Database

Published: 21 June 2023| Version 1 | DOI: 10.17632/jxwvdwhpc2.1
Contributor:
Mayke Pereira

Description

This dataset was created to be used in machine learning tasks for both binary and multiclass classification problems. It consists of 7,909 microscopic images of breast tumor tissue, captured using different magnification factors (40X, 100X, 200X, and 400X). The images in this dataset were resized to 224x224 pixels and organized according to binary and multiclass classification tasks.

Files

Steps to reproduce

The images were acquired, resized in 224x224 and stored in a folder structure, taking into account the binary and multiclass classification tasks, as well as the available magnifications.

Institutions

Instituto Federal de Educacao Ciencia e Tecnologia do Espirito Santo

Categories

Breast Cancer, Machine Learning, Histopathology, Mathematical Morphology

Licence