MR-to-US Registration Using Multiclass Segmentation of Hepatic Vasculature with a Reduced 3D U-Net

Abstract

Accurate hepatic vessel segmentation and registration using ultrasound (US) can contribute to beneficial navigation during hepatic surgery. However, it is challenging due to noise and speckle in US imaging and liver deformation. Therefore, a workflow is developed using a reduced 3D U-Net for segmentation, followed by non-rigid coherent point drift (CPD) registration. By means of electromagnetically tracked US, 61 3D volumes were acquired during surgery. Dice scores of 0.77, 0.65 and 0.66 were achieved for segmentation of all vasculature, hepatic vein and portal vein respectively. This compares to inter-observer variabilities of 0.85, 0.88 and 0.74 respectively. Target registration error at a tumor lesion of interest was lower (7.1 mm) when registration is performed either on the hepatic or the portal vein, compared to using all vasculature (8.9 mm). Using clinical data, we developed a workflow consisting of multi-class segmentation combined with selective non-rigid registration that leads to sufficient accuracy for integration in computer assisted liver surgery.

Registration of Untracked 2D Laparoscopic Ultrasound to CT Images of the Liver Using Multi-Labelled Content-Based Image Retrieval

“…In the case of false positive mis-segmentations, the precision values do not decrease as much, indicating that a future automatic segmentation method should prioritise sensitivity over specificity. For future work, Deep Learning frameworks for segmentation of portal and hepatic veins in liver US images [32] could be integrated.…”

“…Vessel sections are manually segmented and labelled as hepatic vein or portal vein in each LUS image. A future piece of work will consider automating this process, using methods such as the one in [32]. We perform two sets of experiments to assess separately the two components of our registration framework, the CBIR system and the HMM based registration.…”

Computational Methods for Liver Vessel Segmentation in Medical Imaging: A Review

“…The authors of [ 26 ] also employed a deep learning method utilizing the 3D U-Net architecture for the intraoperative segmentation of liver vasculature. Vessels were segmented based on ultrasound images.…”

“…Of the 31 articles included in the literature review: 21 studies presented methods using CT images [ 12 , 13 , 15 , 16 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 27 , 28 , 29 , 30 , 31 , 34 , 35 , 36 , 37 , 38 ] 8 papers concerned approaches that use CTA images [ 7 , 8 , 10 , 11 , 14 , 17 , 18 , 24 ] 2 articles presented methods that use MR images [ 13 , 33 ] 2 studies described approaches using USG images [ 26 , 32 ] 13 papers presented results of research based on publicly available datasets [ 12 , 13 , 15 , 16 , 24 , 25 , 27 , 28 , 29 , 30 , 31 , 37 , 38 ] 18 articles describe results obtained using private datasets [ ...…”

“…18 articles describe results obtained using private datasets [ 7 , 8 , 9 , 10 , 11 , 14 , 17 , 18 , 19 , 20 , 21 , 22 , 26 , 32 , 33 , 34 , 35 , 36 ]…”