Pre-operative planning of multiple probes in three dimensions for liver cryosurgery: comparison of different optimization methods

摘要

Pre-operative planning of percutaneous thermal ablations is a difficult but decisive task for a safe and successful intervention. The purpose of our research is to assist surgeons in preparing cryoablation with an automatic pre-operative path planning algorithm able to propose a placement for multiple needles in three dimensions. The aim is to optimize the tumor coverage problem while taking into account a precise computation of the frozen area. Using an implementation of the precise estimation of the ice balls, this study focuses on the optimization in an acceptable time of multiple probes positions with 5 degrees of freedom, regarding the constraint of optimal volumetric coverage of the tumor by the combined necrosis. Pennes equation was used to solve the propagation of cold within the tissues, and included in an objective function of the optimization process. The propagation computation being time-consuming, seven optimization algorithms from the literature were experimented under different conditions and compared, in order to reduce overall computation time while preserving precision. Moreover, several hybrid algorithms were tested to reduce required time for the computations. Some of these methods were found suitable for the conditions of our cryosurgery planning. We conclude that this combination of bioheat simulation and optimization can be appropriate for a use by practitioners in acceptable conditions of time and precision. Copyright (c) 2015 John Wiley & Sons, Ltd.