Predicting high-intensity focused ultrasound thalamotomy lesions using 2D magnetic resonance thermometry and 3D Gaussian modeling

Evaluation of Cerebral Volume Changes in Patients with Tremor Treated by MRgFUS Thalamotomy

The purpose of the study is to quantify volumetric variations of cortical and subcortical brain structures after Vim ablation using MRgFUS, and correlate them with the patients’ clinical features and treatment outcomes. For this pilot retrospective study we enrolled 31 patients with a mean age of 70.86 years who were eligible for unilateral Vim thalamotomy. Clinical evaluation included tremor severity assessment using the FTM scale and cognitive assessment using the MoCA score. MRI data were acquired with a 3T scanner, using a dedicated 32-channel coil and acquiring a volumetric sequence of T1 3D IR FSPGR (BRAVO), before treatment and one year after MRgFUS thalamotomy. Image processing and volume data extraction were conducted with dedicated software. A volumetric analysis showed a significant reduction (p < 0.05) of the left thalamus 1 year after the treatment in patients with ET. Other significant results were found on the same side in the other nuclei of the basal ganglia and in the cerebellar cortex. In confronting the two groups (ET, PD), no significant differences were found in terms of age, FTM, MoCA scores, or brain volumes. Similarly, no significant correlations were found between the FTM and MoCA scores and the brain volumes before the treatment.

Lesion location and lesion creation affect outcomes after focused ultrasound thalamotomy

MRI-guided focused ultrasound thalamotomy has been shown to be an effective treatment for medication refractory essential tremor. Here, we report a clinical-radiological analysis of 123 cases of MRI-guided focused ultrasound thalamotomy, and explore the relationships between treatment parameters, lesion characteristics and outcomes. All patients undergoing focused ultrasound thalamotomy by a single surgeon were included. The procedure was performed as previously described, and patients were followed for up to 1 year. MRI was performed 24 h post-treatment, and lesion locations and volumes were calculated. We retrospectively evaluated 118 essential tremor patients and five tremor-dominant Parkinson's disease patients who underwent thalamotomy. At 24 h post-procedure, tremor abated completely in the treated hand in 81 essential tremor patients. Imbalance, sensory disturbances and dysarthria were the most frequent acute adverse events. Patients with any adverse event had significantly larger lesions, while inferolateral lesion margins were associated with a higher incidence of motor-related adverse events. Twenty-three lesions were identified with irregular tails, often extending into the internal capsule; 22 of these patients experienced at least one adverse event. Treatment parameters and lesion characteristics changed with increasing surgeon experience. In later cases, treatments used higher maximum power (normalized to skull density ratio), accelerated more quickly to high power, and delivered energy over fewer sonications. Larger lesions were correlated with a rapid rise in both power delivery and temperature, while increased oedema was associated with rapid rise in temperature and the maximum power delivered. Total energy and total power did not significantly affect lesion size. A support vector regression was trained to predict lesion size and confirmed the most valuable predictors of increased lesion size as higher maximum power, rapid rise to high-power delivery, and rapid rise to high tissue temperatures. These findings may relate to a decrease in the energy efficiency of the treatment, potentially due to changes in acoustic properties of skull and tissue at higher powers and temperatures. We report the largest single surgeon series of focused ultrasound thalamotomy to date, demonstrating tremor relief and adverse events consistent with reported literature. Lesion location and volume impacted adverse events, and an irregular lesion tail was strongly associated with adverse events. High-power delivery early in the treatment course, rapid temperature rise, and maximum power were dominant predictors of lesion volume, while total power, total energy, maximum energy and maximum temperature did not improve prediction of lesion volume. These findings have critical implications for treatment planning in future patients.

Comprehensive Evaluation of Factors Affecting Tremor Relapse after MRgFUS Thalamotomy: A Case-Control Study

OBJECTIVE

To identify possible relevant factors contributing to tremor relapse after MRgFUS thalamotomy in patients with essential tremor (ET) and Parkinson's disease (PD).

METHODS

We identified patients with tremor relapse from a series of 79 treatments in a single institution. The demographic and clinical characteristics of the study group patients were compared to those of patients who did not relapse in the same follow-up period. Imaging and procedural factors were compared using a control group matched for clinical and demographic characteristics.

RESULTS

Concerning clinical and demographic characteristics, we did not find statistically significant differences in gender and age. Seventy-three percent of patients with tremor relapse were Parkinson's disease patients. Using MRI, we found larger thalamotomy lesions at the 1-year follow-up in the control group with stable outcomes, compared to patients with tremor relapse. In the tractography evaluation, we found a more frequent eccentric position of the DRTt in patients with tremor relapse.

CONCLUSIONS

The most relevant determining factors for tremor relapse after MRgFUS thalamotomy appear to be tremor from Parkinson's disease and inaccurate thalamic targeting. Size of the thalamotomy lesion can also influence the outcome of treatment.

Comparative evaluation of tractography-based direct targeting and atlas-based indirect targeting of the ventral intermediate (Vim) nucleus in MRgFUS thalamotomy

To analyze and compare direct and indirect targeting of the Vim for MRgFUS thalamotomy. We retrospectively evaluated 21 patients who underwent unilateral MRgFUS Vim ablation and required targeting repositioning during the procedures. For each patient, in the three spatial coordinates, we recorded: (i) indirect coordinates; (ii) the coordinates where we clinically observed tremor reduction during the verification stage sonications; (iii) direct coordinates, measured on the dentatorubrothalamic tract (DRTT) at the after postprocessing of DTI data. The agreement between direct and indirect coordinates compared to clinically effective coordinates was evaluated through the Bland–Altman test and intraclass correlation coefficient. The median absolute percentage error was also calculated. Compared to indirect targeting, direct targeting showed inferior error values on the RL and AP coordinates (0.019 vs. 0.079 and 0.207 vs. 0.221, respectively) and higher error values on the SI coordinates (0.263 vs. 0.021). The agreement between measurements was higher for tractography along the AP and SI planes and lower along the RL planes. Indirect atlas-based targeting represents a valid approach for MRgFUS thalamotomy. The direct tractography approach is a valuable aid in assessing the possible deviation of the error in cases where no immediate clinical response is achieved.

Lesion 3D modeling in transcranial MR-guided focused ultrasound thalamotomy

Transcranial magnetic resonance-guided focused ultrasound (tMRgFUS) allows to perform incisionless thermoablation of deep brain structures. This feature makes it a very useful tool for the treatment of multiple neurological and psychiatric disorders. Currently, feedback of the thermoablation process is based on peak temperature readings assessed on real-time two-dimensional MRI thermometry. However, an accurate methodology relating thermal dosimetry with three-dimensional topography and temporal evolution of the lesion is still to be defined, thus hurdling the establishment of well-defined, evidence-based criteria to perform safe and effective treatments. In here we propose threshold-based thermoablation models to predict the volumetric topography of the lesion (whole lesion and necrotic core) in the short-to-mid-term based on thermal dosimetry estimated from intra-treatment MRI thermometry. To define and validate our models we retrospectively analyzed the data of sixty-three tMRgFUS thalamotomies for treating tremor. We used intra-treatment MRI thermometry to estimate whole-treatment three-dimensional thermal dose maps, defined either as peak temperature reached (T_max) or thermal isoeffective dose (TID). Those maps were thresholded to find the dosimetric level that maximize the agreement (Sorensen-Dice coefficient - SDc) with the boundaries of the whole lesion and its core, assessed on T2w images 1-day (post-24h) and 3-months (post-3M) after treatment. Best predictions were achieved for the whole lesion at post-24h (SDc = 0.71), with T_max /TID over 50.0 °C/90.5 CEM43. The core at post-24h and whole lesion at post-3M lesions reported a similar behavior in terms of shape accuracy (SDc ~0.35), and thermal dose thresholds ~55 °C/4100.0 CEM43. Finally, the optimal levels for post-3M core lesions were 55.5 °C/5800.0 CEM43 (SDc = 0.21). These thermoablation models could contribute to the real-time decision-making process and improve the outcome of tMRgFUS interventions both in terms of safety and efficacy.

MRI-based thermal dosimetry based on single-slice imaging during focused ultrasound thalamotomy

Transcranial MRI-guided focused ultrasound (MRgFUS) is a noninvasive thermal ablation method approved for the treatment of essential tremor and tremor-dominant Parkinson's disease. This method uses MR temperature imaging (MRTI) to monitor the treatment. Accurately tracking the accumulated thermal dose is important for both safety and efficacy. Currently, MRTI is obtained in a single plane that varies between sonications, preventing direct tracking of the accumulated dose. In this work, we tested a method to estimate this dose during 120 MRgFUS treatments. This method used the MRTI to create simulated thermal images for sonications when the imaging plane was changed. This approach accurately predicted the lesion shapes. The mean Sørensen-Dice similarity coefficient between the lesion segmentations and dose regions at the 17 cumulative min at 43 °C (CEM43) threshold used by the device software was 0.82 but varied among different treatments (range: 0.34-0.95). Tissue swelling appeared to explain when mismatch occurred, although other errors probably contributed. Overall, the mean distance between the lesion segmentations and the 17 CEM43 dose contours was 0.37 ± 0.57 mm. The probability for thermal damage was estimated to be 50% at 13.6 CEM43 and a maximum temperature of 48.6 °C. Due to large thermal gradients, which exceeded 99 CEM43/mm on average, the area where the probability for thermal damage was uncertain was narrow. Overall these results show that the 17 CEM43 threshold is on average a good predictor for thermal lesions, although there will always be a narrow margin where the fate of the tissue is uncertain.

Recent technological advancements in thermometry

Thermometry is the key factor for achieving successful thermal therapy. Although invasive thermometry with a probe has been used for more than four decades, this method can only detect the local temperature within the probing volume. Noninvasive temperature imaging using a tomographic technique is ideal for monitoring hot-spot formation in the human body. Among various techniques, such as X-ray computed tomography, microwave tomography, echo sonography, and magnetic resonance (MR) imaging, the proton resonance frequency shift method of MR thermometry is the only method currently available for clinical practice because its temperature sensitivity is consistent in most aqueous tissues and can be easily observed using common clinical scanners. New techniques are being proposed to improve the robustness of this method against tissue motion. MR techniques for fat thermometry were also developed based on relaxation times. One of the latest non-MR techniques to attract attention is photoacoustic imaging.