The ability to automatically learn task specific feature representations has led to a huge success of deep learning methods. When large training data is scarce, such as in medical imaging problems, transfer learning has been very effective. In this paper, we systematically investigate the process of transferring a Convolutional Neural Network, trained on ImageNet images to perform image classification, to kidney detection problem in ultrasound images. We study how the detection performance depends on the extent of transfer. We show that a transferred and tuned CNN can outperform a state-of-the-art feature engineered pipeline and a hybridization of these two techniques achieves 20\% higher performance. We also investigate how the evolution of intermediate response images from our network. Finally, we compare these responses to state-of-the-art image processing filters in order to gain greater insight into how transfer learning is able to effectively manage widely varying imaging regimes.
Submitted 20 Apr 2017 to Computer Vision and Pattern Recognition
Published 21 Apr 2017
Author comments: Published in MICCAI Workshop on Deep Learning in Medical Image Analysis, 2016http://arxiv.org/abs/1704.06040http://arxiv.org/pdf/1704.06040.pdf