# Balanced-MixUp for Highly Imbalanced Medical Image Classification

Source: [Balanced-MixUp for Highly Imbalanced Medical Image Classification](<https://www.sota2.com/research/paper/68fa6d661e209131cea3984c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-04T05:07:41.009Z

Authors: Adrian Galdran, Gustavo Carneiro, Miguel A. Gonz\\'alez Ballester

Published: 2021-09-20T21:31:31.000Z

## Abstract

Highly imbalanced datasets are ubiquitous in medical image classification problems. In such problems, it is often the case that rare classes associated to less prevalent diseases are severely under-represented in labeled databases, typically resulting in poor performance of machine learning algorithms due to overfitting in the learning process. In this paper, we propose a novel mechanism for sampling training data based on the popular MixUp regularization technique, which we refer to as Balanced-MixUp. In short, Balanced-MixUp simultaneously performs regular (i.e., instance-based) and balanced (i.e., class-based) sampling of the training data. The resulting two sets of samples are then mixed-up to create a more balanced training distribution from which a neural network can effectively learn without incurring in heavily under-fitting the minority classes. We experiment with a highly imbalanced dataset of retinal images (55K samples, 5 classes) and a long-tail dataset of gastro-intestinal video frames (10K images, 23 classes), using two CNNs of varying representation capabilities. Experimental results demonstrate that applying Balanced-MixUp outperforms other conventional sampling schemes and loss functions specifically designed to deal with imbalanced data. Code is released at https://github.com/agaldran/balanced\_mixup .

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [DR grading on Eyepacs (test)](<https://www.sota2.com/research/sota/dr-grading-on-eyepacs-test.md>): DR Grading; Eyepacs (test); Quadratic Weighted Kappa: 80.78.

- [DR grading on Messidor-2 (test)](<https://www.sota2.com/research/sota/dr-grading-on-messidor-2-test.md>): DR Grading; Messidor-2 (test); Quadratic Kappa: 85.14.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2109.09850>)

[PDF](<https://arxiv.org/pdf/2109.09850.pdf>)

[Implementation](<https://github.com/agaldran/balanced_mixup>)