Predictive Value of Skull-Related Parameters on Maximum Temperature During MRgFUS in Parkinson's Disease and Essential Tremor

Motivation: The mean value of the skull density ratio (SDR_mean) is not sufficient to accurately predict the maximum temperature (Tmax) during magnetic resonance-guided focused ultrasound (MRgFUS). Goal(s): This study aimed to evaluate the predictive value of skull-related parameters on Tmax and to examine the distribution and characteristics of SDR in the general population. Approach: Multiple machine learning algorithms were applied to investigate whether Tmax could be predicted from combinations of important skull-related parameters. Results: The combination of important skull-related parameters can accurately predict Tmax, particularly when considering the metrics of the proportion of SDR_local < 0.4 and thickness_mean (SDR_local < 0.4). Impact: In this study, we found that indicators of local skull characteristics were able to predict maximum temperature more accurately, which contributes to preoperative patient screening.