Meeting Banner
Abstract #1292

The substantial influence of negative sampling and prevalence when presenting classification results: case study with TOF-MRA

Tommaso Di Noto1, Guillaume Marie1, Sebastien Tourbier1, Guillaume Saliou1, Meritxell Bach Cuadra1,2,3, Patric Hagmann1, and Jonas Richiardi1,4
1Faculty of Biology and Medicine, Department of Radiology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland, 2Medical Image Analysis Laboratory (MIAL), Centre d’Imagerie BioMédicale (CIBM), Lausanne, Switzerland, 3Signal Processing Laboratory (LTS 5), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 4Advanced Clinical Imaging Technology, Siemens Healthcare, Lausanne, Switzerland

One recurrent problem for applying deep learning models in medical imaging is the reduced availability of labelled training data. A common approach is therefore to focus on image patches rather than whole volumes, thus increasing the number of samples. However, for many diseases anomalous patches (positive samples) are outnumbered by negative patches showing no anomaly. Here, we explore different strategies for negative sampling in the context of brain aneurysm detection. We show that classification performances can vary drastically with respect to negative sampling, and that real-world disease or anomaly prevalence can further degrade performance estimates.

This abstract and the presentation materials are available to members only; a login is required.

Join Here