Learning based fiducial driven registration (LEFIR): Evaluating laser ablation changes for neurological applications

In this work, we present a novel learning based fiducial driven registration (LeFiR) scheme. We also investigate a key problem concerning the nature of landmark choices in relation to different aspects of the deformation, such as force direction, magnitude of displacement, deformation location, and native imaging artifacts of noise and intensity non-uniformity. In this work we focus on the problem of attempting to identify the optimal configuration of landmarks for recovering deformation between a target and a moving image via a thin-plate spline (TPS) based registration scheme. Additionally, we employ the LeFiR scheme to model the localized nature of deformation introduced by a new treatment modality - laser induced interstitial thermal therapy (LITT) for treating neurological disorders. Magnetic resonance guided LITT has recently emerged as a minimally invasive alternative to craniotomy for local treatment of brain diseases (such as glioblastoma multiforme (GBM), epilepsy). There is thus a need to understand (in terms of imaging features) the precise changes in the target region of interest following LITT. In order to evaluate LeFiR, we tested the scheme on a synthetic brain dataset and in two real clinical scenarios for treating GBM and epilepsy with LITT. In all cases LeFiR was found to outperform a uniform landmark based TPS registration scheme.