An adaptive targeting algorithm for magnetic resonance-guided high-intensity focused ultrasound controlled hyperthermia.

BACKGROUND: Mild hyperthermia has been demonstrated to improve the efficacy of chemotherapy, radiation, and immunotherapy in various cancer types. One localized, non-invasive method of administering mild hyperthermia is magnetic resonance-guided high-intensity focused ultrasound (MRgHIFU). However, challenges for ultrasound such as beam deflection, refraction and coupling issues may result in a misalignment of the HIFU focus and the tumor during hyperthermia. Currently, the best option is to stop the treatment, wait for the tissue to cool, and redo the treatment planning before restarting the hyperthermia. This current workflow is both time-consuming and unreliable. PURPOSE: An adaptive targeting algorithm was developed for MRgHIFU controlled hyperthermia treatments for cancer therapeutics. This algorithm executes in real time while hyperthermia is being administered to ensure that the focus is within our target region. If a mistarget is detected, the HIFU system will electronically steer the focus of the HIFU beam to the correct target. The goal of this study was to quantify the accuracy and precision of the adaptive targeting algorithm's ability to correct a purposely misplanned hyperthermia treatment in real-time using a clinical MRgHIFU system. METHODS: A gelatin phantom with acoustic properties matched to the average speed of sound in human tissue was used to test the adaptive targeting algorithm's accuracy and precision. The target was purposely offset 10 mm away from the focus at the origin, in four orthogonal directions, allowing the algorithm to correct for this mistarget. In each direction, 10 data sets were collected for a total sample size of 40. Hyperthermia was administered with a target temperature set at 42°C. The adaptive targeting algorithm was run during the hyperthermia treatment and 20 thermometry images were collected after the beam steering occurred. The location of the focus was quantified by calculating the center of heating on the MR thermometry data. RESULTS: The average calculated trajectory passed to the HIFU system was 9.7 mm ± 0.4 mm where the target trajectory was 10 mm. The accuracy of the adaptive targeting algorithm after the beam steering correction was 0.9 mm and the precision was 1.6 mm. CONCLUSION: The adaptive targeting algorithm was implemented successfully and was able to correct the 10 mm mistargets with high accuracy and precision in gelatin phantoms. The results demonstrate the capability to correct the MRgHIFU focus location during controlled hyperthermia.

A robotic magnetic resonance-guided high-intensity focused ultrasound platform for neonatal neurosurgery: Assessment of targeting accuracy and precision in a brain phantom.

BACKGROUND: Intraventricular hemorrhage (IVH) is one of the most serious neurovascular complications resulting from premature birth. It can result in clotting of blood within the ventricles, which causes a buildup of cerebrospinal fluid that can lead to posthemorrhagic ventricular dilation and posthemorrhagic hydrocephalus. Currently, there are no direct treatments for these blood clots as the standard of care is invasive surgery to insert a shunt. Magnetic resonance-guided high-intensity focused ultrasound (MRgHIFU) has been investigated as a noninvasive treatment to lyse blood clots. However, current MRgHIFU systems are not suitable in the context of treating IVH in neonates. PURPOSE: We have developed a robotic MRgHIFU neurosurgical platform designed to treat the neonatal brain. This platform facilitates ergonomic patient positioning and directs treatment through their open anterior fontanelle while providing a larger treatment volume. The platform is based on an MR-compatible robot developed by our group. Further development of the platform has warranted investigation of its targeting ability to assess its feasibility in the neonatal brain. This study aimed to quantify the platform's targeting accuracy, precision, and repeatability using a brain phantom and clinical MRI system. METHODS: A thermosensitive brain-mimicking phantom was developed to test the platform's targeting accuracy. Rectangular grid patterns were created with HIFU thermal energy "lesions" in the phantoms by targeting specific coordinate points. The intended target locations were demarcated by inserting carbon fiber rods through a targeting assessment template. Coordinates for the intended and actual targets were derived from T2-weighted MRI scans, and the centroid distance between them was measured. Subsequently, the platform's targeting accuracy was quantified according to equations derived from ISO Standard 9283:1998. RESULTS: HIFU ablation resulted in distinct thermal lesions within the thermosensitive phantoms, which appeared as discrete hypointense regions in T2-weighted MR scans. A total of 127 target points were included in the data analysis, which yielded a targeting accuracy of 0.6 mm and targeting precision of 1.2 mm. CONCLUSIONS: The robotic MRgHIFU platform was shown to have a high degree of accuracy, precision, and repeatability. The results demonstrate the platform's functionality when targeting through simulated brain matter. These results serve as an initial verification of the platform targeting ability and showed promise toward the final application in a neonatal brain.