Meeting Banner
Abstract #2894

Towards improving high-grade gliomas diagnostic surveillance on T2-weighted images using weak labels from radiology reports

Tommaso Di Noto1, Chirine Atat1, Eduardo Gamito Teiga1, Monika Hegi2, Andreas Hottinger3, Patric Hagmann1, Meritxell Bach Cuadra1, and Jonas Richiardi1
1Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 2Neuroscience Research Center, CHUV, Lausanne, Switzerland, 3Department of Clinical Neurosciences, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland


Manual annotations are a major bottleneck in supervised machine learning. We present a method that leverages Natural Language Processing (NLP) to generate automatic weak labels from radiology reports. We show how weak labels can be used for the image classification task of high-grade-glioma diagnostic surveillance. We apply a convolutional neural network (CNN) to classify T2w difference maps that either indicate tumor stability or instability. Results suggest that pretraining the CNN with weak labels and fine-tuning it on manually-annotated data leads to better performance (though not statistically significant) when compared to a baseline pipeline where only manually annotated data is used.

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

Join Here