Work

Lesion characterization in magnetic resonance-guided focused ultrasound for essential tremor using quantitative susceptibility mapping

Year: 2026

Type: article

Abstract: BACKGROUND: Focused ultrasound (FUS) thalamotomy targeting the ventral intermediate nucleus (VIM) is an increasingly adopted treatment for essential tremor. Although conventional T1/T2-weighted imaging is routinely used for post-treatment lesion segmentation shortly after the intervention, these modalities can be confounded by transient edema. Consequently, alternative sequences may offer refined tissue-level specificity regarding cytotoxic necrosis. OBJECTIVES: We investigated the use of quantitative susceptibility mapping (QSM) as a tool to characterize the thermal lesion and explored the association between QSM-derived measures and clinical outcomes. METHODS: Fifteen essential tremor patients had clinical evaluation, T1, T2-FLAIR, QSM and diffusion images acquisition one week before and three months after FUS-thalamotomy. We assessed clinical improvement using the Clinical Rating Scale for Tremor, alongside sensory and cerebellar adverse effects. Lesions were segmented on individual QSM maps to extract volume and mean magnetic susceptibility. Using automatic region segmentation of the VIM and probabilistic tractography, we reconstructed the dentato-rubro-thalamo-cortical tract and the medial lemniscus, quantifying their spatial overlaps and adverse effect risks. RESULTS: QSM revealed lesions in 100% of cases, compared to 47% and 60% for T1- and T2-weighted images respectively. Mean magnetic susceptibility correlated with clinical improvement. Sensory adverse effect was related to lesion overlap with medial lemniscus, whereas cerebellar symptoms mapped to regions receiving cerebello-thalamic fibers. CONCLUSIONS: QSM is a reliable method for lesion segmentation. Magnetic susceptibility reflects the microstructural changes and tissue damage produced by FUS thalamotomy, serving as an important determinant of therapeutic efficacy.

Source: NeuroImage Clinical

DOI: 10.1016/j.nicl.2026.104083