Improving Pathological Structure Segmentation via Transfer Learning Across Diseases

Barleen Kaur, Paul Lemaı̂tre, Raghav Mehta, Nazanin Mohammadi Sepahvand, Doina Precup, Douglas Arnold, Tal Arbel

Abstract:

One of the biggest challenges in developing robust machine
learning techniques for medical image analysis is the lack of access to
large-scale annotated image datasets needed for supervised learning.
When the task is to segment pathological structures (e.g. lesions, tumors)
from patient images, training on a dataset with few samples is very chal-
lenging due to the large class imbalance and inter-subject variability.
In this paper, we explore how to best leverage a segmentation model
that has been pre-trained on a large dataset of patients images with one
disease in order to successfully train a deep learning pathology segmenta-
tion model for a different disease, for which only a relatively small patient
dataset is available. Specifically, we train a UNet model on a large-scale,
proprietary, multi-center, multi-scanner Multiple Sclerosis (MS) clinical
trial dataset containing over 3500 multi-modal MRI samples with expert-
derived lesion labels. We explore several transfer learning approaches to
leverage the learned MS model for the task of multi-class brain tumor seg-
mentation on the BraTS 2018 dataset. Our results indicate that adapting
and fine-tuning the encoder and decoder of the network trained on the
larger MS dataset leads to improvement in brain tumor segmentation
when few instances are available. This type of transfer learning out-
performs training and testing the network on the BraTS dataset from
scratch as well as several other transfer learning approaches, particularly
when only a small subset of the dataset is available.