AUTHOR=Tardy Mickael , Mateus Diana 

TITLE=Leveraging Multi-Task Learning to Cope With Poor and Missing Labels of Mammograms

JOURNAL=Frontiers in Radiology

VOLUME=Volume 1 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/radiology/articles/10.3389/fradi.2021.796078

DOI=10.3389/fradi.2021.796078

ISSN=2673-8740

ABSTRACT=In breast cancer screening, binary classification of mammograms is a common task aiming to determine whether a case is malignant or benign. 
A Computer-Aided Diagnosis (CADx) system based on a trainable classifier requires clean data and labels coming from a confirmed diagnosis. 
Unfortunately, such labels are not easy to obtain in clinical practice, since the histopathological reports of biopsy may not be available alongside mammograms, while normal cases may not have an explicit follow-up confirmation.
Such ambiguities result either in reducing the number of samples eligible for training or in a label uncertainty that may decrease the performances. In this work, we maximize the number of samples for training relying on multi-task learning. We design a deep-neural-network-based classifier yielding multiple outputs in one forward pass. The predicted classes include binary malignancy, cancer probability estimation, breast density, and image laterality. 
Since few samples  have all classes available and confirmed, we propose to introduce the uncertainty related to the classes as a per-sample weight during training. Such weighting prevents updating the network when training on uncertain or missing labels.
We evaluate our approach on the public INBreast and a private datasets, showing the statistically significant improvements compared to the baseline and an independent state-of-the-art approaches . Moreover, we use mammograms from Susan G. Komen Tissue Bank for fine-tuning, further demonstrating the ability to improve the performances in multi-task learning setup and to deal with raw clinical data.