Optimizing Trajectories for Cranial Laser Interstitial Thermal Therapy Using Computer-Assisted Planning: A Machine Learning Approach

Abstract

Laser interstitial thermal therapy (LITT) is an alternative to open surgery for drug-resistant focal mesial temporal lobe epilepsy (MTLE). Studies suggest maximal ablation of the mesial hippocampal head and amygdalohippocampal complex (AHC) improves seizure freedom rates while better neuropsychological outcomes are associated with sparing of the parahippocampal gyrus (PHG). Optimal trajectories avoid sulci and CSF cavities and maximize distance from vasculature. Computer-assisted planning (CAP) improves these metrics, but the combination of entry and target zones has yet to be determined to maximize ablation of the AHC while sparing the PHG. We apply a machine learning approach to predict entry and target parameters and utilize these for CAP. Ten patients with hippocampal sclerosis were identified from a prospectively managed database. CAP LITT trajectories were generated using entry regions that include the inferior occipital, middle occipital, inferior temporal, and middle temporal gyri. Target points were varied by sequential AHC erosions and transformations of the centroid of the amygdala. A total of 7600 trajectories were generated, and ablation volumes of the AHC and PHG were calculated. Two machine learning approaches (random forest and linear regression) were investigated to predict composite ablation scores and determine entry and target point combinations that maximize ablation of the AHC while sparing the PHG. Random forest and linear regression predictions had a high correlation with the calculated values in the test set (ρ = 0.7) for both methods. Maximal composite ablation scores were associated with entry points around the junction of the inferior occipital, middle occipital, and middle temporal gyri. The optimal target point was the anteromesial amygdala. These parameters were then used with CAP to generate clinically feasible trajectories that optimize safety metrics. Machine learning techniques accurately predict composite ablation score. Prospective studies are required to determine if this improves seizure-free outcome while reducing neuropsychological morbidity following LITT for MTLE.

Multicenter validation of automated trajectories for selective laser amygdalohippocampectomy

“…In addition, trajectory 3 was used in our initial comparative study and was derived from expert consensus. By comparison, trajectory 4 entry was defined solely through machine‐learning parameters . Taking into account the trajectory safety metrics, ablation volume estimations, and external expert feasibility ratings, we would advocate providing the surgeon with a choice of trajectories 3 and 4 for use in future prospective validation studies.…”

“…Automated trajectory algorithms optimize planning parameters in an objective and systematic fashion based on user‐defined parameters . The EpiNav platform is based on the current literature to date and therefore benefits from the combined learning curves of multiple centers as well as the incorporation of machine learning parameters . As further data are acquired, the algorithm is adaptable to continually incorporate and optimized these features.…”

“…Computer‐assisted planning (CAP) for automated generation of LITT trajectories using EpiNav was performed prior to assessment of the manually implemented trajectory and has been described previously. We refer readers to our previous work for a more in‐depth description of the CAP algorithm . In brief, a preoperative T1 image is used to generate a patient‐specific whole‐brain parcellation and pseudoCT using geodesic information flows (GIFs).…”

“…The second incorporates the trajectory parameters derived from expert consensus and incorporates an entry point through the inferior occipital gyrus targeting a 3‐mm anterior, medial, and inferior translation of the centroid of the amygdala (denoted as Trajectory 3 ). Finally, the third trajectory employs trajectory parameters derived from machine learning . These include an entry region at the temporooccipital junction, targeting a 3‐mm anterior and medial translation of the centroid of the amygdala without an inferior translation (denoted as Trajectory 4 ).…”

“…We demonstrated improved trajectory metrics including estimated ablation volumes of the amygdalohippocampal complex (AHC) over manually planned trajectories. We next applied a machine learning approach to deriving the LITT trajectory parameters that further increases the estimated ablation of the AHC, while sparing ablation of the parahippocampal gyrus (PHG) . This revealed that optimal entry points cluster around the temporo‐occipital junction and optimal target points are at the anterior medial aspect of the amygdala.…”