UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Interactive Medical Image Segmentation using Deep Learning with Image-specific Fine-tuning

Wang, G; Li, W; Zuluaga, MA; Pratt, R; Patel, PA; Aertsen, M; Doel, T; ... Vercauteren, T; + view all (2018) Interactive Medical Image Segmentation using Deep Learning with Image-specific Fine-tuning. IEEE Transactions on Medical Imaging , 37 (7) pp. 1562-1573. 10.1109/TMI.2018.2791721. (In press). Green open access

[thumbnail of David VoR08270673.pdf]
Preview
Text
David VoR08270673.pdf - Published Version

Download (4MB) | Preview

Abstract

Convolutional neural networks (CNNs) have achieved state-of-the-art performance for automatic medical image segmentation. However, they have not demonstrated sufficiently accurate and robust results for clinical use. In addition, they are limited by the lack of image-specific adaptation and the lack of generalizability to previously unseen object classes. To address these problems, we propose a novel deep learning-based framework for interactive segmentation by incorporating CNNs into a bounding box and scribble-based segmentation pipeline. We propose image-specific fine-tuning to make a CNN model adaptive to a specific test image, which can be either unsupervised (without additional user interactions) or supervised (with additional scribbles). We also propose a weighted loss function considering network and interaction-based uncertainty for the fine-tuning. We applied this framework to two applications: 2D segmentation of multiple organs from fetal MR slices, where only two types of these organs were annotated for training; and 3D segmentation of brain tumor core (excluding edema) and whole brain tumor (including edema) from different MR sequences, where only tumor cores in one MR sequence were annotated for training. Experimental results show that 1) our model is more robust to segment previously unseen objects than state-of-the-art CNNs; 2) image-specific fine-tuning with the proposed weighted loss function significantly improves segmentation accuracy; and 3) our method leads to accurate results with fewer user interactions and less user time than traditional interactive segmentation methods.

Type: Article
Title: Interactive Medical Image Segmentation using Deep Learning with Image-specific Fine-tuning
Open access status: An open access version is available from UCL Discovery
DOI: 10.1109/TMI.2018.2791721
Publisher version: http://dx.doi.org/10.1109/TMI.2018.2791721
Language: English
Additional information: This is an open access article published under the terms of the IEEE OA licence. For information on re-use, please refer to the publisher’s terms and conditions.
Keywords: Interactive image segmentation, Convolutional neural network, Fine-tuning, Fetal MRI, Brain tumor
UCL classification: UCL
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL EGA Institute for Womens Health
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL EGA Institute for Womens Health > Maternal and Fetal Medicine
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Med Phys and Biomedical Eng
URI: https://discovery.ucl.ac.uk/id/eprint/10032237
Downloads since deposit
808Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item