Share Email Print
cover

Proceedings Paper • new

Deep residual networks for automatic segmentation of laparoscopic videos of the liver
Author(s): Eli Gibson; Maria R. Robu; Stephen Thompson; P. Eddie Edwards; Crispin Schneider; Kurinchi Gurusamy; Brian Davidson; David J. Hawkes; Dean C. Barratt; Matthew J. Clarkson
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Motivation: For primary and metastatic liver cancer patients undergoing liver resection, a laparoscopic approach can reduce recovery times and morbidity while offering equivalent curative results; however, only about 10% of tumours reside in anatomical locations that are currently accessible for laparoscopic resection. Augmenting laparoscopic video with registered vascular anatomical models from pre-procedure imaging could support using laparoscopy in a wider population. Segmentation of liver tissue on laparoscopic video supports the robust registration of anatomical liver models by filtering out false anatomical correspondences between pre-procedure and intra-procedure images. In this paper, we present a convolutional neural network (CNN) approach to liver segmentation in laparoscopic liver procedure videos. Method: We defined a CNN architecture comprising fully-convolutional deep residual networks with multi-resolution loss functions. The CNN was trained in a leave-one-patient-out cross-validation on 2050 video frames from 6 liver resections and 7 laparoscopic staging procedures, and evaluated using the Dice score. Results: The CNN yielded segmentations with Dice scores ≥0.95 for the majority of images; however, the inter-patient variability in median Dice score was substantial. Four failure modes were identified from low scoring segmentations: minimal visible liver tissue, inter-patient variability in liver appearance, automatic exposure correction, and pathological liver tissue that mimics non-liver tissue appearance. Conclusion: CNNs offer a feasible approach for accurately segmenting liver from other anatomy on laparoscopic video, but additional data or computational advances are necessary to address challenges due to the high inter-patient variability in liver appearance.

Paper Details

Date Published: 3 March 2017
PDF: 6 pages
Proc. SPIE 10135, Medical Imaging 2017: Image-Guided Procedures, Robotic Interventions, and Modeling, 101351M (3 March 2017); doi: 10.1117/12.2255975
Show Author Affiliations
Eli Gibson, Univ. College London (United Kingdom)
Radboud Univ. Medical Ctr. (Netherlands)
Maria R. Robu, Univ. College London (United Kingdom)
Stephen Thompson, Univ. College London (United Kingdom)
P. Eddie Edwards, Univ. College London (United Kingdom)
Crispin Schneider, The Royal Free Hospital (United Kingdom)
Kurinchi Gurusamy, The Royal Free Hospital (United Kingdom)
Brian Davidson, The Royal Free Hospital (United Kingdom)
David J. Hawkes, Univ. College London (United Kingdom)
Dean C. Barratt, Univ. College London (United Kingdom)
Matthew J. Clarkson, Univ. College London (United Kingdom)


Published in SPIE Proceedings Vol. 10135:
Medical Imaging 2017: Image-Guided Procedures, Robotic Interventions, and Modeling
Robert J. Webster; Baowei Fei, Editor(s)

Video Presentation

Deep-residual-networks-for-automatic-segmentation-of-laparoscopic-videos-of



© SPIE. Terms of Use
Back to Top