Robust ultrasound probe tracking: initial clinical experiences during robot-assisted partial nephrectomy

Philip Pratt,Alexander Jaeger,Archie Hughes-Hallett,Erik Mayer,Justin Vale,Ara Darzi,Terry Peters,Guang-Zhong Yang
DOI: https://doi.org/10.1007/s11548-015-1279-x
2015-01-01
International Journal of Computer Assisted Radiology and Surgery
Abstract:Purpose In order to assist in the identification of renal vasculature and tumour boundaries in robot-assisted partial nephrectomy, robust ultrasound probe calibration and tracking methods are introduced. Contemporaneous image guidance during these crucial stages of the procedure should ultimately lead to improved safety and quality of outcome for the patient, through reduced positive margin rates, segmental clamping, shorter ischaemic times and nephron-sparing resection. Methods Small KeyDot markers with circular dot patterns are attached to a miniature pickup ultrasound probe. Generic probe calibration is superseded by a more robust scheme based on a sequence of physical transducer measurements. Motion prediction combined with a reduced region-of-interest in the endoscopic video feed facilitates real-time tracking and registration performance at full HD resolutions. Results Quantitative analysis confirms that circular dot patterns result in an improved translational and rotational working envelope, in comparison with the previous chessboard pattern implementation. Furthermore, increased robustness is observed with respect to prevailing illumination levels and out-of-focus images due to relatively small endoscopic depths of field. Conclusion Circular dot patterns should be employed in this context as they result in improved performance and robustness. This facilitates clinical usage and interpretation of the combined video and ultrasound overlay. The efficacy of the overall system is demonstrated in the first human clinical case.
What problem does this paper attempt to address?