Share Email Print
cover

Proceedings Paper

Method for evaluation of predictive models of microwave ablation via post-procedural clinical imaging
Author(s): Jarrod A. Collins; Daniel Brown; T. Peter Kingham; William R. Jarnagin; Michael I. Miga; Logan W. Clements
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Development of a clinically accurate predictive model of microwave ablation (MWA) procedures would represent a significant advancement and facilitate an implementation of patient-specific treatment planning to achieve optimal probe placement and ablation outcomes. While studies have been performed to evaluate predictive models of MWA, the ability to quantify the performance of predictive models via clinical data has been limited to comparing geometric measurements of the predicted and actual ablation zones. The accuracy of placement, as determined by the degree of spatial overlap between ablation zones, has not been achieved. In order to overcome this limitation, a method of evaluation is proposed where the actual location of the MWA antenna is tracked and recorded during the procedure via a surgical navigation system. Predictive models of the MWA are then computed using the known position of the antenna within the preoperative image space. Two different predictive MWA models were used for the preliminary evaluation of the proposed method: (1) a geometric model based on the labeling associated with the ablation antenna and (2) a 3-D finite element method based computational model of MWA using COMSOL. Given the follow-up tomographic images that are acquired at approximately 30 days after the procedure, a 3-D surface model of the necrotic zone was generated to represent the true ablation zone. A quantification of the overlap between the predicted ablation zones and the true ablation zone was performed after a rigid registration was computed between the pre- and post-procedural tomograms. While both model show significant overlap with the true ablation zone, these preliminary results suggest a slightly higher degree of overlap with the geometric model.

Paper Details

Date Published: 18 March 2015
PDF: 7 pages
Proc. SPIE 9415, Medical Imaging 2015: Image-Guided Procedures, Robotic Interventions, and Modeling, 94152F (18 March 2015); doi: 10.1117/12.2082910
Show Author Affiliations
Jarrod A. Collins, Vanderbilt Univ. (United States)
Daniel Brown, Vanderbilt Univ. Medical Ctr. (United States)
T. Peter Kingham, Memorial Sloan-Kettering Cancer Ctr. (United States)
William R. Jarnagin, Memorial Sloan-Kettering Cancer Ctr. (United States)
Michael I. Miga, Vanderbilt Univ. (United States)
Vanderbilt Univ. Medical Ctr. (United States)
Logan W. Clements, Vanderbilt Univ. (United States)


Published in SPIE Proceedings Vol. 9415:
Medical Imaging 2015: Image-Guided Procedures, Robotic Interventions, and Modeling
Robert J. Webster; Ziv R. Yaniv, Editor(s)

© SPIE. Terms of Use
Back to Top