Ensembles of Multiple Models and Architectures for Robust Brain Tumour Segmentation

TL;DR

This paper introduces EMMA, an ensemble approach combining multiple models and architectures to improve brain tumor segmentation robustness, reducing overfitting and achieving top performance in the BRATS 2017 challenge.

Contribution

The paper presents EMMA, a novel ensemble method that aggregates diverse models to enhance segmentation accuracy and robustness in brain tumor analysis.

Findings

EMMA outperforms individual models in segmentation accuracy.

EMMA achieved first place in the BRATS 2017 competition.

Ensembling reduces overfitting and dependency on specific model parameters.

Abstract

Deep learning approaches such as convolutional neural nets have consistently outperformed previous methods on challenging tasks such as dense, semantic segmentation. However, the various proposed networks perform differently, with behaviour largely influenced by architectural choices and training settings. This paper explores Ensembles of Multiple Models and Architectures (EMMA) for robust performance through aggregation of predictions from a wide range of methods. The approach reduces the influence of the meta-parameters of individual models and the risk of overfitting the configuration to a particular database. EMMA can be seen as an unbiased, generic deep learning model which is shown to yield excellent performance, winning the first position in the BRATS 2017 competition among 50+ participating teams.