A deep learning approach for 2D ultrasound and 3D CT/MR image registration in liver tumor ablation. (July 2021)

• Record Type: Journal Article
• Authors: WEI, Wei; Haishan, Xu; Alpers, Julian; Rak, Marko; Hansen, Christian
• Is Part Of: Computer methods and programs in biomedicine. Volume 206(2021)
• Publication Date: 2021-07
• Subjects: Slice-to-volume -- Ultrasound -- Image registration -- Image classification -- Image segmentation; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2021.106117
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: A novel pipeline for 2D Ultrasound image to 3D CT/MR volume registration. Use classification network to predict slice to volume orientation. Employ segmentation network to solve registration problem. A distance map based DICE loss function is proposed for the registration tasks. Abstract: Background and Objective: Liver tumor ablation is often guided by ultrasound (US). Due to poor image quality, intraoperative US is fused with preoperative computed tomography or magnetic tomography (CT/MR) images to provide visual guidance. As of today, the underlying 2D US to 3D CT/MR registration problem remains a very challenging task. Methods: We propose a novel pipeline to address this registration problem. Contrary to previous work, we do not formulate the problem as a regression task, which - for the given registration problem - achieves a low performance regarding accuracy and robustness due to the limited US soft-tissue contrast and the inter-patient variability on liver vessels. Instead, we first estimate the US probe angle roughly by using a classification network. Given this coarse initialization, we then improve the registration by formulation of the problem as a segmentation task, estimating the US plane in the 3D CT/MR through segmentation. Results: We benchmark our approach on 1035 clinical images from 52 patients, yielding average registration errors of 11.6 ° and 4.7 mm, which outperforms the state of the art SVR method[1]. Conclusion: Our results show the efficiency of the proposed registration pipeline, which has potential to improve the robustness and accuracy of intraoperative patient registration.