Near real-time system for high-resolution computationalTMS navigation

用于高分辨率计算 TMS 导航的近实时系统

• 批准号: 10558627
• 负责人: Aapo Nummenmaa
• 金额: $ 76.85万
• 依托单位: MASSACHUSETTS GENERAL HOSPITAL
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-02-01 至 2026-11-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10558627
• 关键词: 3-Dimensional; Acceleration; Anatomy; Applications Grants; Area; Benchmarking; Boundary Elements; Brain; Clinical; Clinical Research; Computing Methodologies; Data; Development; Dose; Electromagnetics; Electromyography; Functional Magnetic Resonance Imaging; Generations; Geometry; Goals; Head; Hot Spot; Human; Human Volunteers; Individual; Individual Differences; Location; Magnetic Resonance Imaging; Major Depressive Disorder; Manufacturer; Maps; Measures; Memory; Methods; Modeling; Motor; National Institute of Mental Health; Neuronavigation; Neurosciences Research; Optics; Patient Schedules; Patients; Performance; Peripheral; Physiologic pulse; Positioning Attribute; Prefrontal Cortex; Publishing; Real-Time Systems; Research; Resolution; Scalp structure; Skin; Source; Specific qualifier value; Speed; Surface; System; Techniques; Technology; Testing; Therapeutic; Time; Tissues; Transcranial magnetic stimulation; automated segmentation; clinical application; clinical effect; commercial application; cortex mapping; cranium; electric field; gray matter; healthy volunteer; improved; innovation; magnetic field; neurosurgery; novel; operation; physical model; prototype; repetitive transcranial magnetic stimulation; response; translational potential; treatment-resistant depression; volunteer; white matter

Project Summary/Abstract: Transcranial Magnetic Stimulation (TMS) is a widely used technology for non-invasive modulation of human brain activity. TMS induces electric fields (E-fields) in the intracranial tissue by means of time-varying magnetic fields (electromagnetic induction) that results in the possibility of obtaining suprathreshold stimulation intensities safely and with little discomfort to the subjects. Clinical applications of TMS include major depressive disorder (MDD) and treatment resistant depression (TRD) in which repetitive TMS (rTMS) is administered to Dorsolateral Prefrontal Cortex (DLPFC) with well-demonstrated efficacy. Both in clinical and basic neuroscience research applications, it is important that the stimulation is accurately targeted to the desired region(s). The E-field distribution that is induced by the TMS pulse in the intracranial tissue is the key physical quantity that can be used to delineate which areas are stimulated and which are not. This is especially important for non-motor regions such as the DLPFC because a direct peripheral measure (e.g., electromyographic response) cannot be used to guide the stimulation. To date, computationally estimated E- field distributions have been used in “online” commercial neuronavigation systems to guide the TMS coil positioning, but the currently available systems offer only spherically symmetric head models that cannot properly take into account the individual differences in tissue geometries and may result in substantial targeting and dosing errors. On the other hand, the most accurate computational methods for E-field estimation are too slow to enable near real time operation. Therefore, no technique exists that has the computational efficiency to enable neuronavigation applications while at the same time incorporating high level of anatomical detail and numerical accuracy. To remove this efficiency vs. accuracy dilemma that is currently posing a critical barrier for development of more quantitative TMS approaches, we propose to use our recently developed Boundary Element Method (BEM) based computational strategy accelerated by the Fast Multipole Method (FMM) that is suitable for both online and offline application scenarios. Our approach starts with developing an automatic segmentation and surface generation pipeline to obtain accurate representations of the tissue conductivity boundaries using individual MRI data. We will subsequently develop and experimental TMS neuronavigation system that utilizes the fast BEM-FMM method. The purpose of this system is to render the E-field distributions on top of the 3D brain anatomy and to guide the operator to position the TMS coil and associated E-field “hot spot” to the desired location. We will interface the computational engine with a commercial TMS navigator to demonstrate translational potential for clinical research and ultimately to therapeutic/clinical applications. Finally, we will validate the computational neuronavigation accuracy with an anthropomorphic head phantom and evaluate our system in healthy volunteers and in volunteer neurosurgical patients that are scheduled for invasive direct cortical mapping that will be used as a gold standard.

项目摘要/摘要： 经颅磁刺激(TMS)是一种应用广泛的非侵入性人体调节技术 大脑活动。TMS通过时变磁场在颅内组织中产生电场 导致获得阈值以上刺激可能性的场(电磁感应) 强度安全，对受试者几乎没有不适。TMS的临床应用主要包括 抑郁症(MDD)和难治性抑郁症(TRD)，其中重复TMS(RTMS) 应用于背外侧前额叶皮质(DLPFC)，效果良好。无论是在临床上还是在 在基础神经科学研究应用中，重要的是刺激要准确地针对 所需区域(S)。TMS脉冲在颅内组织中产生的电场分布是关键 可用于描述哪些区域受到刺激，哪些区域不受刺激的物理量。这是特别的 对于诸如DLPFC的非运动区很重要，因为直接的外周测量(例如， 肌电反应)不能用来指导刺激。到目前为止，通过计算估计的E- 场分布已被用于在线商业神经导航系统中以引导TMS线圈 定位，但目前可用的系统只提供球对称头部模型，不能 适当考虑组织几何形状的个体差异，并可能导致大量靶向 以及给药错误。另一方面，电场估计最准确的计算方法也是 启用近乎实时操作的速度较慢。因此，不存在具有计算效率的技术来 支持神经导航应用，同时结合高水平的解剖细节和 数值精度。要消除效率与准确性的两难困境，目前这一难题正在构成 为了开发更定量的TMS方法，我们建议使用我们最近开发的边界 基于单元方法(BEM)的计算策略由快速多极子方法(FMM)加速 适用于线上和线下应用场景。我们的方法从开发一种自动的 分割和表面生成管道以获得组织电导率的准确表示 使用单个核磁共振数据的边界。我们随后将开发和实验TMS神经导航 使用快速边界元-FMM方法的系统。该系统的目的是渲染E场分布 在3D脑部解剖之上，并指导操作员将TMS线圈和相关的电磁场定位为热的 Spot“到所需的位置。我们将把计算引擎与商业TMS导航器对接，以 展示临床研究的转化潜力，并最终转化为治疗/临床应用。 最后，我们将使用一个拟人化的头部模型来验证计算神经导航的准确性 并在健康志愿者和自愿神经外科患者中评估我们的系统 将被用作黄金标准的侵入性直接皮质标测。