Acoustic Modeling of skull bone for improved transcranial MR-guided focused ultrasound therapy

项目摘要/摘要(限行30行) 该建议旨在改进特定于患者的计算机体层摄影(CT)派生的建模 通过将声学和热模拟与水听器扫描进行比较来实现经颅聚焦超声 自发性震颤切除颅骨瓣及临床磁共振测温(MRTI) 丘脑切开术治疗。磁共振引导经颅聚焦超声(TMRgFUS)是一种非 用于治疗各种神经疾病的侵入性治疗方法。TMRgFUS紧密依赖于 聚焦超声束穿过不均匀的人类头骨。一个根本性的挑战是准确地 确定颅骨的声学特性以对不均匀进行相位补偿。此外， 诸如声速C和衰减α的声学参数可以随着温度的升高而改变， 导致进一步的散焦。不准确的声学参数可能导致偏离目标的加热、更长的治疗时间 次数和失败的治疗。 这个项目将通过准确地确定人体头骨来提高超声波的聚焦能力。 个体头骨声学参数。混合角谱(HAS)波束模拟方法及 Pennes生物热方程可以通过映射声学和热学来模拟压力场和热升 CT Hounsfield单位的参数。这些模拟的结果可以与实验数据进行比较 确定tFUS声学和热学建模的准确性。反过来应用此方法，代理 优化算法以其在黑箱代价优化问题中的优势，将被用来迭代求解 使用成本函数调整模拟参数以符合实验数据。Aim我将确定这一关系 向CT Hounsfield单元提供骨的声学特性。优化算法将迭代地调整 声学参数映射，使得比较模拟和测量的发射声学的成本函数 压力是最小的。由此得到的最佳声学参数准确地模拟了经颅声学 变速箱。AIM II将确定高声功率处理效率降低的原因 TMRgFUS.优化算法将反复调整声学和热学参数以最小化 成本函数模拟与临床ET丘脑切除术患者MRTI数据的比较。 这项工作将改进通过人类头骨进行的声学建模，这是改进的第一步 经颅聚焦超声治疗。根据聚焦超声基金会的说法，tMRgFUS可能是 适用于至少34种神经性疾病。因此，这项工作可能会产生显著的放大效应 降低神经学领域的发病率和死亡率。

Project Summary/Abstract (Limit 30 lines) This proposal aims to improve the patient-specific Computed Tomography (CT)-derived modeling of transcranial focused ultrasound by comparing acoustic and thermal simulations to hydrophone scans of excised skull flaps and clinical magnetic resonance thermometry (MRTI) from Essential Tremor (ET) thalamotomy treatments. Magnetic resonance-guided transcranial focused ultrasound (tMRgFUS) is a non- invasive therapeutic modality used to treat a wide variety of neurological disorders. tMRgFUS relies on tightly focusing the ultrasound beam through the inhomogeneous human skull. A fundamental challenge is accurately determining the acoustic properties of the skull to phase-compensate for the inhomogeneities. Furthermore, acoustic parameters such as speed of sound c and attenuation α may change with increased temperature, causing further defocusing. Inaccurate acoustic parameters can result in off-target heating, longer treatment times, and failed treatments. This project will improve the focusing of ultrasound through the human skull by accurately determining individual skull acoustic parameters. The Hybrid Angular Spectrum (HAS) beam simulation method and the Pennes bioheat equation can simulate pressure fields and thermal rises by mapping acoustic and thermal parameters to CT Hounsfield Units. The results of these simulations may be compared to experimental data to determine the accuracy of tFUS acoustic and thermal modeling. Applying this method in reverse, a surrogate optimization algorithm, which excels at black-box expensive optimization problems, will be used to iteratively adjust simulation parameters to fit experimental data using a cost function. Aim I will determine the relationship of the acoustic properties of bone to CT Hounsfield Units. An optimization algorithm will iteratively adjust the acoustic parameter mapping such that a cost function comparing simulated and measured transmitted acoustic pressures is minimized. The resulting optimal acoustic parameters accurately model transcranial acoustic transmission. Aim II will determine the cause of reduced treatment efficiency with high acoustic powers during tMRgFUS. An optimization algorithm will iteratively adjust the acoustic and thermal parameters to minimize a cost function comparing simulation to MRTI data from clinical ET Thalamotomy patients. This work will improve acoustic modeling through the human skull, which is the first step in improving transcranial focused ultrasound therapy. According to the Focused Ultrasound Foundation, tMRgFUS could be applied to at least 34 neurological disorders. Thus, this work could have a magnified effect, significantly reducing morbidity and mortality across the field of neurology.