Casey Halpern Uses Epilepsy Technique to Map Human Cravings

Casey Halpern利用癫痫技术绘制人类渴求图谱 脑深部电刺激无法普及治疗饱受肥胖和精神危机困扰的5000万美国人，促使神经外科医生将侵入性电极标定作为过渡手段，争分夺秒研发可规模化的非侵入性疗法。 本期节目深入探讨了立体定向功能神经外科，详细介绍了侵入性脑深部电刺激（DBS）和非侵入性经颅聚焦超声如何用于调节大脑的奖赏与运动回路。Halpern博士解释了强迫行为和渴求背后的具体解剖学结构，指出功能亢进的皮层区域和伏隔核是强迫症、成瘾及进食障碍的核心驱动因素。他重点介绍了利用立体定向脑电图（sEEG）等癫痫类标定技术来分离症状特异性神经特征的最新临床突破。讨论最终强调，尽管侵入性手术能为严重的难治性患者带来改变人生的缓解效果，但其真正价值在于为面向更广泛人群开发可规模化、非侵入性的人工智能与技术疗法提供必要的基础图谱。 虽然美国约90%的普通神经外科手术侧重于脊柱治疗，但立体定向功能神经外科则专注于脑深部电刺激（DBS）和非侵入性经颅聚焦超声等局部靶向手段。 (0:00) 脑深部电刺激本质上是一种精准的硬件传递工具，用于将靶向电流输送至运动回路等特定大脑区域深处，其本身并非一种独立的疗法。 (0:00) 帕金森病治疗过程中的偶然观察发现，对边缘回路进行电刺激可使强迫行为和重度抑郁等临床精神症状消失。 (2:19) 强迫特质存在于一个谱系上，对追求卓越的专业人士而言可以成为宝贵财富，但当其失控超出个人的主观意愿时，就会演变为使人衰弱的临床疾病。 (4:19) 约30%的强迫症患者对标准药物治疗或专业的暴露与反应预防（ERP）行为疗法没有反应。 (8:15)

Dr. Casey Halpern argues that resolving severe psychiatric and compulsive disorders requires direct surgical mapping of the brain's corticostriatal reward circuits. Because invasive neurosurgery cannot scale to treat population-wide mental health crises, intracranial electrophysiology must serve as a temporary blueprint to guide targeted, non-invasive neuromodulation. Neural Circuitry of Compulsion and Impulsivity (10:12) According to Dr. Halpern, compulsive behaviors stem from dysregulation within corticostriatal circuits. Hyperfunctioning cortical areas, specifically the prefrontal and orbitofrontal cortices, project into subcortical structures of the basal ganglia, including the caudate, putamen, and dorsal striatum. These pathways connect directly to the ventral striatum, where the nucleus accumbens serves as a critical gatekeeper for reward-seeking behavior. When this gating mechanism fails, individuals pursue rewards despite severe negative consequences—a pathological feature shared across clinical obsessive-compulsive disorder (OCD), binge eating disorder, bulimia, and chemical addictions. Halpern distinguishes compulsion from impulsivity. Compulsion is defined as the inability to resist an urge despite knowing the associated risks, such as an OCD patient checking locks until 3:00 a.m. or a shock-conditioned rodent continuing to seek food. Impulsivity, by contrast, is characterized by a failure of patience and an inability to delay gratification, such as failing to wait for a cue before seeking a reward. Both behavioral traits exist on a spectrum, operating as adaptive traits in high-demand professions when regulated, but becoming debilitating disorders when cognitive control collapses. Conventional Treatment Limits and Current Surgical Efficacy (8:15) First-line psychiatric interventions—such as exposure response prevention (ERP) and serotonin-modulating medications like SSRIs or tricyclic antidepressants—leave approximately 30% of OCD patients without relief. Dr. Halpern notes that cognitive behavioral therapy (CBT) often yields only temporary benefits, with high relapse rates once therapy ends. In laboratory assessments using eye-tracking hardware, severe patients aware of observation through one-way mirrors still proceeded to binge following negative mood provocations, demonstrating that conscious awareness alone cannot override biological loss of control. For patients with treatment-resistant conditions, neurosurgical options include thermal capsulotomy—which permanently ablates a 3 to 4 mm section of brain tissue—and deep brain stimulation (DBS), which delivers targeted electrical currents via implanted micro-wires. While brain tissue displays notable resilience following these interventions, clinical efficacy remains constrained. Halpern notes that current surgical responder rates sit at approximately 50%, with responders still experiencing residual symptoms. This modest success rate drives the need for more symptom-specific circuit mapping. Translating Electrophysiology to Track Cravings and Obsessions (12:34) To enhance surgical precision, Halpern adapts intraoperative methods from Parkinson's disease surgeries, where surgeons convert cellular electrical activity into audio signals to locate 'tremor cells' firing at the exact frequency of a physical tremor. Halpern identifies 'craving' as the subjective biological equivalent of a tremor for eating disorders and addictions. In a single-case study involving an awake OCD patient, researchers successfully isolated obsession-related cells intraoperatively, establishing proof of concept for real-time symptom mapping. Because therapeutic home-use implants utilize 1.0 mm electrodes that capture aggregate signals from thousands of neurons—compared to the 0.1 mm single-cell wires used in acute surgery—researchers reproduce symptoms in laboratory settings to map circuit dynamics. Psychiatrists employ validated mood provocations to trigger the specific emotional distress that precedes a patient's binge. While binge eating targets were translated directly from preclinical mouse models, Halpern emphasizes that complex conditions like OCD and depression cannot be modeled reliably in animals and require direct human electrophysiological mapping. Invasive Mapping to Resolve the Scalability Bottleneck (28:38) To safely isolate pathological circuits, neurosurgeons utilize stereoencephalography (sEEG), an invasive monitoring method using sub-millimeter depth electrodes adapted from epilepsy management. Halpern highlights that reversible electrodes offer a safer exploratory platform than irreversible ablation, noting ongoing collaborative work at Baylor College of Medicine and UCSF to map depressive circuits. However, surgical interventions face an absolute scalability bottleneck: conditions like obesity, addiction, and depression impact over 50 million Americans, whereas only approximately 200,000 DBS procedures have been performed globally throughout medical history. Halpern maintains that invasive neurosurgery must primarily serve to inform non-invasive, scalable treatments. Modalities like transcranial magnetic stimulation (TMS) already hold FDA approvals for depression, OCD, and smoking cessation, while transcranial MR-guided focused ultrasound offers incision-free ablation for essential tremor and Parkinson's disease. By confirming targets invasively first, researchers aim to optimize focused ultrasound and non-invasive screening models—including external biometric algorithms, such as acoustic vocal tracking developed at the University of Washington—to anticipate behavioral relapses before conscious onset. Although neurotechnologies and machine learning models present substantial opportunities for psychiatric care, Halpern emphasizes that the fundamental mechanisms governing deep brain stimulation and non-invasive neuromodulation remain incompletely understood. Direct recording inside the living human brain remains an essential preliminary step to establish reproducible anatomical targets before scalable, non-invasive therapies can reliably treat population-scale compulsive disorders.

神经外科专家Casey Halpern指出，强迫与冲动行为源于大脑皮层与边缘环路（特别是伏隔核）的门控机制失常，而脑深部电刺激（DBS）等侵入性技术正用于绘制精准神经图谱，为未来开发可规模化的非侵入性疗法奠定基础。 强迫行为的环路机制与伏隔核门控 (10:12) Halpern将强迫症（OCD）置于一个广泛的光谱中，指出适度的细致与执念常是外科医生或高管的职业优势，但失控时则演变为严重障碍。在病理状态下，患者的前额叶与眶额皮层表现出持续的过度活跃，并向下投射至基底节（包括尾状核、壳核及背侧纹状体）与腹侧纹状体。腹侧纹状体中的伏隔核作为奖赏寻求的核心门控枢纽，一旦该区域机制失灵，就会驱动个体无视严重后果去满足冲动，例如强迫症患者检查门锁至凌晨3点、暴食症患者失控进食，或成瘾者接触危险毒贩。Halpern区分了强迫与冲动的生物学本质：强迫是不顾风险屈从于冲动，而冲动则是缺乏耐心和行为抑制能力。 难治性强迫症的手术疗效与局限 (8:15) 约30%的强迫症患者对一线药物（如靶向血清素系统的SSRI或三环类抗抑郁药）及暴露反应预防（ERP）行为疗法无反应。对于这些难治性患者，神经外科主要提供脑深部电刺激（DBS）与内囊切开术（capsulotomy）。内囊切开术通过热能永久损毁3至4毫米的脑组织，而DBS则通过植入带有微型触点的细绝缘导线传递靶向电流。Halpern指出，DBS作为一种可逆的电调控手段，在定位不理想时可关闭或取出，因而被部分学者认为安全性更高。然而，目前这两种外科手段的临床反应率仅约为50%，且应答患者仍残留部分症状。Halpern本人每年仅开展约3至5例此类手术，这促使研究团队致力于寻找更精准的症状特异性靶点。 借鉴帕金森病与癫痫的术中定位技术 (12:34) 为了在脑内定位病理性冲动，Halpern团队借鉴了帕金森病手术的经验。帕金森病手术通过微电极将丘脑底核的细胞电活动转化为音频信号，通过寻找与手部震颤频率一致的“震颤细胞”来精确定位。团队将这一逻辑迁移至强迫症与进食障碍，试图捕捉与主观“渴求”（craving）或执念相匹配的神经元活动，并已在单例清醒强迫症手术中成功记录到执念相关细胞。此外，临床上正在引入癫痫领域成熟的立体定向脑电图（sEEG）技术，利用直径小于1毫米的深部电极进行监测与刺激，以绘制抑郁和强迫症的异常环路。Baylor医学院与UCSF的团队已率先将sEEG与精神病学结合用于严重抑郁的研究，Halpern团队则在等待FDA批准将该侵入性绘图用于强迫症。 诱发实验与治疗耐药背后的生理失控 (24:33) 对于极度耐药的患者而言，单纯的认知提升无法阻止行为失控，许多患者在完全清楚后果的情况下依然无法抑制冲动。为了捕捉暴食发作前的特征性脑电信号，研究人员将植入长期治疗设备的患者置于实验室中，通过专业的情绪诱发方案激发其个人的负面情绪诱因，并在单向玻璃与眼动仪监测下记录患者的颅内电活动。长期植入设备的电极直径约为1毫米，记录的是数千个神经元的综合信号，而非急性术中所用的0.1毫米单细胞电极。实验显示，即使明确知晓处于科学监测中，重度患者仍会发生暴食。这印证了认知行为疗法（CBT）在严重器质性失控面前的局限性，其效果常在疗程结束后消退，因而必须通过生物学手段恢复其行为控制力。 非侵入性技术探索与公共卫生规模化缺口 (28:38) 经颅磁刺激（TMS）目前已获FDA批准用于抑郁、强迫症和尼古丁成瘾，临床常将其作为侵入性手术前的筛选工具；经颅磁共振引导聚焦超声（MRgFUS）亦获批用于帕金森病及特发性震颤的无切口消融，且正在试验用于暂时性开放血脑屏障以递送肿瘤药物。然而，Halpern强调，无论是超声还是TMS，要拓展至精神疾病和肥胖症，核心瓶颈在于缺乏精准解剖坐标。更为严峻的是规模化难题：肥胖、阿片类药物成瘾、抑郁和自杀倾向等危机影响着超过5000万美国人，而全球迄今累计完成的DBS手术仅约20万例。Halpern认为，神经外科必须先通过侵入性手段在活体人脑中绘制精准坐标，才能指导人工智能和可穿戴监测（如华盛顿大学利用声音和呼吸特征预测抑郁发作），最终反哺开发出安全、非侵入且可大众普及的治疗方案。 侵入性神经外科手术虽能为极少数极重度精神与冲动障碍患者提供治疗线索，但因操作复杂、成本高昂且目前临床有效率仅维持在50%左右，无法独立应对千万人规模的公共卫生危机。当前的核心挑战在于如何在严格的科学假说与转化研究驱动下，利用脑内精准图谱攻克靶点未知的瓶颈，从而引导开发出真正具备循证依据的非侵入性调控技术。

## Background 该内容由 Andrew Huberman 于 2026 年 5 月 7 日发布在 YouTube 平台的 Huberman Lab Essentials 节目中。视频简介指出，本期嘉宾为宾夕法尼亚大学医学院神经外科教授 Casey Halpern 医生，双方深入探讨了如何利用脑深部电刺激（DBS）等神经调控手段治疗帕金森病、强迫症（OCD）、暴食症及抑郁症，并剖析了驱动冲动与渴望的大脑回路及前沿的无创预测与治疗工具。 ## TL;DR 本期访谈围绕神经外科在精神和行为障碍治疗中的前沿应用展开。Halpern 医生详细解析了伏隔核等核心脑回路如何被强迫行为和成瘾“绑架”，导致患者不顾风险一意孤行。对话探讨了从侵入式的脑深部电刺激（DBS）、立体脑电图（sEEG）到无创的经颅磁刺激（TMS）和聚焦超声技术的临床现状与突破。面对庞大的公共卫生危机，专家指出未来必须结合人工智能与机器学习，开发出更具扩展性的普适医疗解决方案。 ## Atmosphere & style 这场对话兼具严谨的科学洞察与深厚的人文关怀，整体基调专业、前沿且充满启发性。主持人 Huberman 运用生动的比喻与追问，与 Halpern 医生谦逊、幽默且理性的外科医生视角完美共鸣。随着话题从立竿见影的震颤治疗奇迹，转向难治性精神病患明知失控却无能为力的残酷现实，谈话展现出一种直面医学极限的深刻张力。最终，对未来技术交融的乐观展望与对粗制滥造设备的尖锐批判，使整场访谈在温和融洽的氛围中，传递出推动人类医学走向极端尖端（Extreme cutting edge）的澎湃激情。 ## Key takeaways - [00:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=0s) 神经外科诊疗范围极其广泛，高度细分的立体定向与功能神经外科目前专注于脑深部电刺激术（DBS）和经颅聚焦超声等前沿手段。 - [00:20](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=20s) DBS 的本质是通过植入大脑深部导线末端的微小触点输送微弱电流，如同精准作用于极小区域的“电流药物”。 - [02:18](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=138s) 刺激旨在缓解帕金森震颤的电极由于波及负责情感处理的边缘回路，意外使患者的强迫行为融化消散，为精神疾病治疗带来新启示。 - [04:19](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=259s) 强迫症（OCD）是一种谱系障碍，适度的强迫特质在社会中可以是资产，只有当其失控并带来巨大痛苦时才演变成疾病。 - [08:15](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=495s) 常规心理与药物疗法对约 30% 的中重度 OCD 患者失效，临床上面临手术选择的权衡，且目前手术的临床响应率仅在 50% 左右。 - [10:11](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=611s) 伏隔核是主导奖赏寻求行为的“把关人”，其回路被绑架是强迫症、暴食症和物质成瘾患者“冒着风险也要满足强烈渴望”的共同核心分母。 - [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 在 DBS 手术中，医生可以通过将脑细胞电信号转换为音频，实时聆听与帕金森震颤或暴食“渴求”（Craving）完美同步的脑电声音。 - [16:50](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1010s) 非侵入式的经颅磁刺激（TMS）虽然在精准度上仍显模糊，但可作为前期筛查工具，用以定义回路并评估侵入式手术的适合度。 - [18:54](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1134s) 磁共振引导聚焦超声作为无创消融手段已能立竿见影地治疗震颤，但将其应用于精神疾病的最大瓶颈在于尚未明确具体的消融靶点。 - [21:24](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1284s) 精神医学界正借鉴癫痫治疗中高度安全的立体脑电图（sEEG）技术，试图通过积攒人体大脑数据为抑郁症等疾病绘制清晰的靶点图谱。 - [24:33](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1473s) 重度“治疗抵抗型”暴食症患者在高度清醒且深知处于监控的实验室环境下，一旦情绪被诱发依然会不可控地暴食，揭示了生理驱动力的残酷性。 - [28:38](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1718s) 面对美国高达五千万受肥胖、成瘾等困扰的庞大患病群体，历史上仅开展二十万例的 DBS 手术刀无法根本解决问题，必须借助 AI 寻求普适方案。 - [32:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1920s) 医疗市场需要的是能真正治愈疾病的手段，而非纯粹浪费金钱和时间的花哨可穿戴工具，而人类对机器学习在医学上的利用才刚触及冰山一角。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=0s) 神经外科与神经科学的“航天员” 安德鲁·胡伯曼（Andrew Huberman）教授在《Huberman Lab Essentials》节目中迎来了宾夕法尼亚大学医学院（Penn Medicine）的立体定向与功能神经外科主任凯西·哈尔珀恩（Casey Halpern）医生。胡伯曼形象地将神经外科医生比作“神经科学领域的航天员”。为了让大众分清神经外科、神经内科与精神科的区别，哈尔珀恩阐述了神经外科极其广泛的诊疗范围。全国大约 90% 的神经外科手术其实集中在脊柱问题上，如椎间盘突出和腰椎融合术，此外还包括切除脑瘤、夹闭脑动脉瘤、处理创伤性脑损伤（如脑震荡），甚至治疗腕管综合征等周围神经系统疾病。 随着医学的发展，如今的神经外科已经高度细分。哈尔珀恩医生目前专注于**脑深部电刺激术（Deep Brain Stimulation, DBS）**以及近期获得美国食品药品监督管理局（FDA）批准的**经颅聚焦超声（Transcranial Focused Ultrasound）**。后者作为一种无创的脑部消融（Ablation）手段，目前被批准用于治疗震颤。 - **DBS 手术原理**：将一根绝缘的微细导线植入大脑深部（如与帕金森病相关的区域），其核心治疗手段并非导线本身，而是通过导线末端的多个微小触点输送微弱的电刺激。 - **电刺激的本质**：哈尔珀恩将其幽默地比喻为“植入一种用来输送药物的工具”，只是这里的“药物”是以电流的形式精准作用于大脑的极小区域。 --- ### [02:18](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=138s) 从帕金森震颤到意外的“情绪疗效” 哈尔珀恩医生表示，能够以这种方式与人类大脑互动是一项极大的特权，且始终以提供有意义的治疗为目标。在临床中，当电极对准目标区域时，如果刺激到仅几毫米之隔的邻近区域，可能会引发短暂的副作用，比如突然放声大笑或瞬间恐慌。有趣的是，这些“副作用”往往会转化为新的治疗灵感。 医生们发现，一些患有帕金森病并伴有抑郁症或强迫症（OCD）等精神共病的患者，在接受旨在缓解震颤的 DBS 治疗后，其冲动和强迫行为（如赌博成瘾）竟然“融化消散”了，情绪也大为改善。哈尔珀恩解释说，这不仅是因为震颤消失带来了心理慰藉，更因为电极同时刺激到了非运动回路——即负责情感处理的**边缘回路（Limbic Circuit）**。 - **科学启示**：通过逐步学习如何治疗性地调节这些边缘区域，科学家可以针对抑郁症等其他适应症开发新型疗法。 - **最震撼的临床瞬间**：让饱受 20 年折磨的震颤患者在诊所内接受电刺激后，震颤瞬间得到缓解。正是这种立竿见影的奇迹，激发了哈尔珀恩在大学时期立志成为一名神经外科医生的梦想。 --- ### [04:19](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=259s) 强迫症（OCD）的本质、谱系与现有疗法 在探讨强迫症（OCD）时，哈尔珀恩坦言自己的视角与每天接诊大量患者的精神科医生有所不同。他作为研究型学者和外科医生，每年大约仅为 3 到 5 例强迫症患者实施 DBS 手术，因此其核心实验室专注于如何通过理解大脑回路来阻断“强迫观念（Obsession）”并终止由此产生的“强迫行为（Compulsion）”。目前，他正联合全美多位专家，像通过电极手术寻找癫痫病灶那样，探寻强迫症在大脑中的起源，并结合影像学和遗传学进行全方位研究。 哈尔珀恩将 OCD 视为一种**谱系障碍**。他幽默地自嘲道，作为一名神经外科医生，他对手术安全的执着近乎“强迫观念”，对手术步骤的执行近乎“强迫行为”。在某种程度上，可控的强迫特质在社会中是一种资产，许多著名的 CEO、外科医生和科学家都带有这种特质。只有当这种特质失控、让人无法自拔且带来巨大痛苦时，它才演变成真正的强迫症。 - **一线药物治疗**：针对强迫症，首选药物是**选择性 5-羟色胺再摄取抑制剂（SSRIs）**和**三环类抗抑郁药（Tricyclics）**。由于 5-羟色胺系统会与去甲肾上腺素和多巴胺系统相互交织，因此很难做到绝对的单一系统针对性，这也解释了为何很难预测药物对特定患者的具体疗效。 - **心理与行为疗法**：**暴露反应预防疗法（Exposure Response Prevention, ERP）**被认为是极其有效的认知行为疗法。宾夕法尼亚大学由埃德娜·福阿（Edna Foa）创立的临床机构便专注于此，通过让患者暴露于压力源并逐渐习惯（Habituate），从而减少强迫行为，帮助患者回归日常功能。 --- ### [08:15](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=495s) 难治性 OCD 的手术权衡与大脑消融术 尽管上述疗法极其卓越，但仍有约 30% 的强迫症患者无法从中获益，其中许多人处于中重度到重度状态。面对这些常规疗法失效的患者，哈尔珀恩医生极具救治动力。然而，现有的手术选择虽值得追求，却仍非最优解。目前的主要手术手段包括 **DBS 术**和**内囊切开术（Capsulotomy）**。内囊切开术是一种消融手段，它不留存电极，而是直接通过加热来破坏大脑中的特定一小块组织（大约 3 到 4 毫米大小）。 关于大脑消融，医学界存在不同的声音：有人认为那部分大脑就像“阑尾”一样，切除很安全；有人则认为通过电极进行功能调制（Modulate）更安全。哈尔珀恩指出，大脑的神奇之处有时恰恰在于它的“毫无负面反应”——手术穿过某些脑区时，患者的各项功能测试完全不受影响，小范围消融后除了短暂的术后恢复，也无明显副作用。然而，即便手术足够安全，目前的**临床响应率（Responder Rate）也仅为 50% 左右**，且获得响应的患者依然会残留强迫症状。这种现状激励着哈尔珀恩去寻找更具疾病和症状特异性的精准输送疗法。 --- ### [10:11](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=611s) 大脑核心回路与驱动冒险的“核心冲动” 当被问及若要在强迫症患者脑中进行探测或消融会选择何处时，哈尔珀恩指出 OCD 是一种涉及**皮质（Cortex）与皮质下（Subcortex）**的疾病。患者的皮质区域（如前额叶皮质和眶额皮质）往往处于功能亢进（Hyperfunctioning）状态，需要被正常化；而这些区域会投影到皮质下结构，即基底核（包括尾状核、壳核或背侧纹状体）。这些结构又与**腹侧纹状体（Ventral Striatum）**紧密相连。 哈尔珀恩将大量精力集中在腹侧纹状体，其中包括**伏隔核（Nucleus Accumbens）**。该区域负责控制和“把关”奖赏寻求行为。当该区域受到扰乱时，它会放行强迫性行为。例如，在动物实验中，大鼠会冒着遭受足部电击（Punishment）的风险去追求奖赏；这与强迫症患者的临床表现高度相似。 - **强迫症典型表现**：由于内心的强烈冲动，患者会反复检查房屋安全直至凌晨三点，不惜牺牲整晚的睡眠；明知 dropped 牙刷掉在地上不卫生，却会陷入长达数小时洗手或刷牙的强迫行为中。这种“**不顾风险，一意孤行**”的判断模式是此类疾病的核心。 - **行为成瘾的共性**： * **进食障碍**：暴食症患者会暴饮暴食，催吐症（Bulimia）患者会强行导吐，全然不顾身体受到的危害。 * **物质成瘾**：吸毒成瘾者会为了获取毒品（如海洛因或阿片类药物）向毒贩支付高额费用，不惜面临致命的法律与健康风险。 - **核心结论**：这种“冒着风险也要满足强烈渴望”的冲动，是强迫症、暴食症和成瘾行为的共同分母，也是当今社会最普遍的疾病根源之一，而伏隔核及其接收皮质投影的回路正是主导这一机制的核心网络。 --- ### [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 伏隔核如何被“绑架”以及震颤的脑电声音 伏隔核作为奖赏回路的核心，与大脑诸多区域互联。哈尔珀恩医生阐述道，对奖赏产生渴望是一种正常的生理现象，并非手术要消灭的目标；只有当这种渴望开始将自身或他人置于危险之中时，才需要被干预。重复接触成瘾药物或极强烈的奖赏刺激，在某种程度上会“**绑架（Hijack）**”伏隔核的正常功能。因此，手术的目的是为了打破这种已经固化的习惯或循环发作的病理状态。以严重的暴食症患者为例，他们往往维持着大约每天暴食一次的固定频率。 为了解决这个问题，哈尔珀恩在手术室里借鉴了治疗帕金森病的成熟技术。在为帕金森患者的手术中，当医生将电极准确植入运动结构（如丘脑下核）时，可以通过仪器将脑细胞的电信号转换为音频。 > 胡伯曼好奇地追问：“听起来是不是就像震颤表现出来的样子？” > 哈尔珀恩极为肯定地回答：“完全一样，信号的频率和手抖的频率完美同步。” 医生们在手术中通过极其精细的导线“一次一针（One poke at a time）”地谨慎探测，实时聆听脑电活动的音频，直到捕捉到与震颤频率相同的细胞群，随后通过电刺激使其精准静默，震颤便会瞬间溶解。 - ** appetite 与暴食冲动在 DBS 中的对应物**：哈尔珀恩明确指出，对应帕金森“震颤”的病理标志，在进食障碍和肥胖症中就是**“渴求”（Craving）**。 - **为何选择“Craving”这个词**：因为这个词最能引发患者的共鸣。如果问暴食症或肥胖症患者是否存在“Craving”，他们的答案通常是笃定的“有”；但如果问他们是否“失去了控制”或“在暴食”，他们有时反而无法理解或无法理性察觉到自己的失控。 ### [16:01](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=961s) 强迫症个案研究与术中意识 为了探索大脑中的“渴望”机制，研究团队开展了一项针对强迫症（OCD）患者的单一案例研究。强迫症与成瘾行为密切相关，在大脑的靶点区域也非常相似。在这项研究中，团队尝试在手术室中识别出与“强迫观念”相关的特定细胞，并成功实现了电极位置的优化。这种方法提供了一种概念验证，即只要患者在手术过程中能够耐受清醒状态，医生就能够在手术室内诱发出疾病特异性的症状。 演讲者指出，虽然并非所有患者在进行此类手术时都需要保持清醒，但对于探索大脑特定功能区域的“首次人体临床试验”而言，这种让患者保持清醒并在术中进行交互的方法是极其关键的。 * **研究性质**：针对强迫症患者的单一案例研究（Single case study），旨在优化电极放置。 * **核心突破**：在手术室内成功诱发出疾病特异性的症状，并初步识别出与强迫观念相关的细胞。 --- ### [16:50](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1010s) 经颅磁刺激（TMS）在精神疾病中的应用与局限 对话转入对非侵入式脑刺激技术的探讨。主持人提到，诸如经颅磁刺激（TMS）这类无需穿透颅骨的技术已被用于治疗抑郁症等多种大脑综合征，但其空间精准度似乎相对有限。演讲者对此表示认同，并坦言医学界对包括深部脑刺激（DBS）在内的许多技术的具体工作机制尚未完全明确。这些非侵入式方法因机制不明，在精准度上确实有些“模糊”（fluffy），仍有极大的提升空间。 不过，TMS 已经获得了 FDA 的多项批准。演讲者透露，他们正在积极研究如何利用 TMS 来明确一个特定的大脑回路。如果对该回路进行调节，就能在一定程度上暂时缓解强迫症症状。对于这部分获得暂时性改善的患者，后续便可评估其是否适合接受更具针对性的侵入式深部脑刺激手术。 * **FDA 批准适应症**：TMS 目前已获 FDA 批准用于治疗抑郁症、强迫症（OCD）和尼古丁成瘾。 * **临床转化路径**：利用非侵入式的 TMS 作为前期筛查和回路定义工具，为后续的侵入式手术铺路。 * **研究空白**：在进食障碍（如厌食症和肥胖症）领域，TMS 的相关研究和应用目前还处于极度匮乏的状态。 --- ### [18:54](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1134s) 聚焦超声技术（Focused Ultrasound）的奇迹与未来 磁共振引导聚焦超声（MRI-guided focused ultrasound）是另一种极为前沿的非侵入式技术。目前，它作为一种非侵入式的“消融”（ablation）手段已获 FDA 批准，并在临床上常规用于治疗帕金森病或特发性震颤。演讲者分享了他在临床操作中的震撼感受，将其称之为“奇迹”。因为无需切开皮肤，也无需向大脑深部插入电极，就能在患者的优势手或症状较重的一侧实现立竿见影的治疗效果。 除了直接消融病灶，包括演讲者在内的众多科研人员正致力于将其发展为一种“调制”（modulatory）工具，用于驱动或抑制神经元活动。此外，还有多项临床试验正在探索利用超声波暂时打开血脑屏障，以便将药物精准输送到特定大脑区域（例如脑肿瘤病灶）。 * **临床现状**：已获 FDA 批准用于治疗震颤，无需开颅即可精准消融功能异常的神经区域。 * **前沿探索**： * 尝试利用超声进行神经调制（兴奋或抑制神经活动）。 * 利用超声打开血脑屏障，进行靶向药物输送。 * **核心瓶颈**：尽管该技术在震颤治疗上效果显著，但若想将其应用于精神疾病、肥胖或进食障碍，最大的医学难题在于“不知道具体的消融靶点在哪里”。 --- ### [21:24](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1284s) 癫痫立体脑电图（SEEG）对精神医学的启示 为了找到更精准的治疗靶点，美国正掀起一场向欧洲学习的技术革命——引入立体脑电图（立体定向脑电图，sEEG）。这种方法在癫痫治疗中已非常成熟且高度安全。医生会将直径小于 1 毫米的微细导线深入植入到大脑的各个疑似病灶区。患者随之入院接受监测，以便医生精准捕捉癫痫发作的起始与传播路径。随后，医生可以通过这些电极进行刺激，观察症状变化，进而决定是实施手术切除、激光消融还是植入长期刺激器。 令人兴奋的是，医学界现在对将这种成熟的癫痫定位技术跨界应用于精神健康领域产生了极大的兴趣。演讲者透露，他们正在等待 FDA 的批准，以尝试将该技术用于强迫症患者。同时，他赞扬了贝勒医学院（Baylor）和加州大学旧金山分校（UCSF）的同行，他们已经成功将癫痫的电极植入技术与精神病学相结合，深入研究如何优化抑郁症的电极靶点定位。这种可逆的探索方式（电极可随时关闭或移除）能够积攒宝贵的数据，长远来看，有可能为抑郁症绘制出一条清晰的非侵入式超声消融靶点图谱。 * **技术特点**：使用直径小于 1 毫米的微型微丝电极进行侵入式深部脑电图监测，并发症发生率极低。 * **跨界融合**：将癫痫的成熟诊疗经验引入精神病学，通过大样本的数据积累，最终旨在推导出针对抑郁症等疾病的非侵入式超声治疗靶点。 * **小鼠模型局限性**：肥胖和暴食症的机制可以通过小鼠模型较好地模拟并推导至人体；但抑郁症和强迫症等更为复杂的精神状态极难在小鼠身上复制，因此必须直接在人体大脑中进行深入研究。 --- ### [24:33](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1473s) 治疗抵抗型患者的清醒暴食实验 主持人提到，如果患者能够提高对自身内部状态的感知（Awareness），在产生渴望、暴食冲动甚至自杀念头之前捕捉到这一微弱的信号，或许能成为最强大的自我救赎工具。演讲者非常赞同感知力的重要性，但他随即揭示了一个残酷的临床现实：身为外科医生，他们面对的是那些尝试了认知行为疗法（CBT）、药物治疗和行为管理后依然无效的“治疗抵抗型”（Refractory）重度患者。这些人对自己即将失控的处境心知肚明，有着极高的清醒感知，但生理上的驱动力让他们“依然无法控制自己”。 为了在科学上捕捉这种失控，研究团队在实验室中开展了严谨的实验。他们将带有植入式设备的患者请到实验室，通过一种非常成熟、被称为“情绪诱发”（Mood provocation）的方法，由精神科医生和进食障碍专家现场诱导出每位患者自述的、最容易触发暴食的情绪。 * **实验设计**：在实验室中复制暴食冲动。患者佩戴眼动追踪仪以实时记录其注视焦点和进食动作，同时将录像与大脑内数千个细胞的电信号（通过直径 1 毫米的电极探测）进行精确的时间同步。 * **失控现实**：即便患者坐在单向玻璃后，在充满监控、且高度清醒地知道自己正在接受科学研究的实验室环境下，一旦情绪被诱发，他们依然会不可控地开始暴食。“这正是这种疾病的核心——即便感知清醒，也依然无法控制行为。” * **CBT 的局限**：认知行为疗法（CBT）效果显著，但对重度患者而言其疗效缺乏持续性，一旦停止治疗，许多患者便会迅速倒退回原有的行为模式中。 --- ### [28:38](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1718s) 人工智能、语音监测与公共卫生的尺度 主持人提出，人工智能和机器学习在这一领域大有可为。例如，华盛顿大学的实验室正在通过分析声音特征、呼吸频率和睡眠质量等海量综合线索，在严重抑郁症患者自己产生意识之前，精准预测自杀或抑郁发作的迹象，这就像搜索引擎比我们自己更了解我们的喜好一样。 演讲者对这一方向给予了高度肯定，并指出“我们必须先深入到大脑内部，才能真正走出来”。只有通过侵入式研究搞清楚大脑内部的底层信号，我们才能真正理解并利用好呼吸、声音等外在的非侵入式信号。在这个语境下，他特别厘清了“强迫”（Compulsion，即冒着风险去追求奖赏的冲动）与“冲动”（Impulsivity，如小鼠等不及配对提示音就去抢食行为）的区别。 suicide（自杀）就是其中最危险的一种“冲动”。 最后，面对肥胖、阿片类药物危机、抑郁症和自杀倾向等已然席卷全美乃至全球的流行病级公共卫生危机，演讲者表达了作为一名外科医生的行业局限与对可扩展（Scalable）方案的迫切期盼。 > “我们至今总共也只做过大约 200,000 例深部脑刺激手术。而我们现在讨论的（患病）群体，光是在美国就有 50,000,000 人。这之间根本没办法……” * **冲动与强迫**：强迫（Compulsion）是明知有风险却依然顺从渴望；冲动（Impulsivity）则表现为缺乏耐受力和等待能力的越界行为。 * **严峻的现实数字**： * **全球 DBS 手术总量**：历史上累计仅开展了约 **200,000** 例。 * **美国潜在患者基数**：受肥胖、成瘾、精神疾病困扰的美国民众高达 **50,000,000** 人。面对如此庞大的群体，单一的手术刀无法解决根本问题，医学界必须借助 AI 以及非侵入式设备的辅助，开发出更具扩展性的普适解决方案。 ### [32:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1920s) 神经外科与可穿戴设备的未来展望 演讲者探讨了神经外科医生在解决复杂脑部问题中的潜在角色。尽管目前的医疗技术已经取得了一定进展，但他们更希望能激发一些针对这些疑难杂症的创新倡议，并推动其向更严谨的方向发展。演讲者直言不讳地指出，当下医疗市场最不需要的就是“某种花哨的可穿戴工具，纯粹浪费人们的金钱和时间”。公众和患者需要的是真正能够治愈疾病的疗法。不过，演讲者也澄清自己并非全盘否定当前讨论的这些设备，实际上它们蕴含着巨大的潜力。 在实验室的研究中，机器学习已经成为了必不可少的日常工具。演讲者幽默且谦逊地表示，自己并不是从事这类工作的电气工程师或计算神经科学家，他的核心工作其实是“帮助围绕这些研究提出假说，并协助为其进行资金筹集”。他坚信这一领域未来的广阔前景，但同时也理性地预测，人类目前在如何最有效地利用这些技术方面，仅仅只是“触及了冰山一角”。 - 实验室日常会使用机器学习技术进行相关研究 - 演讲者的主要职责是提出科学假说以及主持团队的筹资（Fundraising）工作 - 呼吁市场减少流于表面、浪费金钱和时间的花哨可穿戴设备，转而专注研发真正的治疗手段（Real therapies） ### [32:44](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1964s) 对医学与神经外科新星的启迪及致谢 主持人对演讲者分享的诸多前沿工具和见解表达了由衷的感谢。主持人满怀期待地推测，屏幕前一定有很多听众在聆听了今天的对话后，会被激发出对医学、尤其是对神经外科的浓厚兴趣。主持人打趣并由衷地感慨，在通往神经外科医生的道路上，“当然必须先成为一名医生”，而今天的对谈极有可能会促成这样一段段美丽的医学职业生涯的开启。 最后，主持人向演讲者致以了崇高的敬意，感谢他能从极其忙碌且重要的日程中抽空前来。主持人试图用“前沿”来形容演讲者的工作，并幽默地自我纠错称，在这里“最好不要用‘喋血前沿（Bleeding edge）’这个词，因为在这个语境下听起来不太对劲”，引发了轻松的氛围。主持人赞扬演讲者的工作正处于人类理解大脑运作及修复方式的“极端尖端（Extreme cutting edge）”。演讲者对此感到非常荣幸，并向主持人及全体听众致谢，访谈在融洽温和的气氛中圆满结束。 - 主持人预测本次对话将激励部分听众投身医学和神经外科领域 - 神经外科医生的培养前提是必须首先成为一名合格的医师（Physician） - 主持人高度评价演讲者的工作处于“极端的剪裁尖端（Extreme cutting edge）”，正带领人类探索大脑修复的终极奥秘 ## Notable quotes > [00:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=0s) 神经科学领域的航天员 — 安德鲁·胡伯曼 > [00:20](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=20s) 植入一种用来输送药物的工具 — 凯西·哈尔珀恩 > [10:11](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=611s) 不顾风险，一意孤行 — 凯西·哈尔珀恩 > [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 绑架（Hijack） — 凯西·哈尔珀恩 > [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 听起来是不是就像震颤表现出来的样子？ — 安德鲁·胡伯曼 > [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 完全一样，信号的频率和手抖的频率完美同步。 — 凯西·哈尔珀恩 > [12:34](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=754s) 一次一针（One poke at a time） — 凯西·哈尔珀恩 > [18:54](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1134s) 不知道具体的消融靶点在哪里 — 凯西·哈尔珀恩 > [24:33](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1473s) 依然无法控制自己 — 凯西·哈尔珀恩 > [24:33](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1473s) 这正是这种疾病的核心——即便感知清醒，也依然无法控制行为。 — 凯西·哈尔珀恩 > [28:38](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1718s) 我们必须先深入到大脑内部，才能真正走出来 — 凯西·哈尔珀恩 > [28:38](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1718s) 我们至今总共也只做过大约 200,000 例深部脑刺激手术。而我们现在讨论的（患病）群体，光是在美国就有 50,000,000 人。这之间根本没办法…… — 凯西·哈尔珀恩 > [32:00](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1920s) 某种花哨的可穿戴工具，纯粹浪费人们的金钱 and 时间 — 凯西·哈尔珀恩 > [32:44](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1964s) 最好不要用‘喋血前沿（Bleeding edge）’这个词，因为在这个语境下听起来不太对劲 — 安德鲁·胡伯曼 > [32:44](https://www.youtube.com/watch?v=aKDzF_fMMrc&t=1964s) 极端的剪裁尖端（Extreme cutting edge） — 安德鲁·胡伯曼 ## Mentioned Andrew Huberman, Casey Halpern, Huberman Lab Essentials, 宾夕法尼亚大学医学院, FDA, Edna Foa, 华盛顿大学, 贝勒医学院, 加州大学旧金山分校 [Source](https://www.youtube.com/watch?v=aKDzF_fMMrc)

[0:00] 安德鲁·胡伯曼: 神经科学领域的航天员 [0:20] 凯西·哈尔珀恩: 植入一种用来输送药物的工具 [10:11] 凯西·哈尔珀恩: 不顾风险，一意孤行 [12:34] 凯西·哈尔珀恩: 绑架（Hijack） [12:34] 安德鲁·胡伯曼: 听起来是不是就像震颤表现出来的样子？ [12:34] 凯西·哈尔珀恩: 完全一样，信号的频率和手抖的频率完美同步。 [12:34] 凯西·哈尔珀恩: 一次一针（One poke at a time） [18:54] 凯西·哈尔珀恩: 不知道具体的消融靶点在哪里 [24:33] 凯西·哈尔珀恩: 依然无法控制自己 [24:33] 凯西·哈尔珀恩: 这正是这种疾病的核心——即便感知清醒，也依然无法控制行为。 [28:38] 凯西·哈尔珀恩: 我们必须先深入到大脑内部，才能真正走出来 [28:38] 凯西·哈尔珀恩: 我们至今总共也只做过大约 200,000 例深部脑刺激手术。而我们现在讨论的（患病）群体，光是在美国就有 50,000,000 人。这之间根本没办法…… [32:00] 凯西·哈尔珀恩: 某种花哨的可穿戴工具，纯粹浪费人们的金钱 and 时间 [32:44] 安德鲁·胡伯曼: 最好不要用‘喋血前沿（Bleeding edge）’这个词，因为在这个语境下听起来不太对劲 [32:44] 安德鲁·胡伯曼: 极端的剪裁尖端（Extreme cutting edge）

[10:12] Dr. Casey Halpern: doing something because of the urge, but despite the risk [4:19] Dr. Casey Halpern: obligated to better understand where the obsessions in the brain come from [4:19] Dr. Casey Halpern: obsessive about safety and compulsive about my surgical procedures [16:01] Dr. Casey Halpern: really critical [16:01] Dr. Casey Halpern: a little fluffy [16:01] Dr. Casey Halpern: probe non-invasively with purpose [18:54] Dr. Casey Halpern: kind of a miracle [24:33] Dr. Casey Halpern: if we can improve awareness, we can improve outcomes [28:38] Dr. Casey Halpern: we have to get in the brain before we get out of it [28:38] Andrew Huberman: I've been that mouse [32:44] Dr. Casey Halpern: wearable fancy tools [32:44] Dr. Casey Halpern: there's lots of promise [32:44] Dr. Casey Halpern: scratching the surface on how best to do it [35:24] Andrew Huberman: certainly you need to be a physician before you can become a neurosurgeon [35:24] Andrew Huberman: bleeding edge cuz in this context it's not going to sound right

[0:00] Welcome to Huberman Lab Essentials, [0:01] [music] where we revisit past episodes [0:04] for the most potent and actionable [0:06] science-based tools for mental health, [0:08] physical health, and performance. [0:11] I'm Andrew Huberman and I'm a professor [0:13] of neurobiology and ophthalmology at [0:15] Stanford School of Medicine. And now for [0:18] my discussion with Dr. Casey Halpern. [0:20] Casey, I should say Dr. Halpern, [0:23] welcome. [0:23] >> Thank you. [0:24] >> Great to be here. [0:24] >> You're a neurosurgeon, which I consider [0:26] the astronauts of neuroscience. For [0:28] those that aren't familiar with the [0:30] differences between neurosurgery, [0:31] neurology, psychiatry, could you just [0:34] educate us a bit? What does a [0:36] neurosurgeon do and how do you think [0:38] about and conceptualize the brain? [0:39] >> Yeah, yeah, the scope of neurosurgery is [0:41] quite broad. We take out brain tumors, [0:43] we clip aneurysms in the brain, we take [0:46] care of patients that have had traumatic [0:47] brain injury, um concussion, uh spine [0:50] surgeries. 90% of what neurosurgeons do [0:53] around the country, uh you know, taking [0:55] care of herniated discs and lumbar [0:57] fusions. So, [1:00] you know, the the scope is [1:02] the entire central nervous system [1:04] includ- including the peripheral nervous [1:05] system. We take care of patients with [1:06] carpal tunnel syndrome and nerve [1:08] disorders. Historically, neurosurgeons [1:11] did everything in that domain, but now [1:14] we subspecialize and I I'm lucky to be [1:16] at Penn Medicine where we can focus on [1:20] one of these areas. So, I'm uh chief of [1:23] stereotactic functional neurosurgery. [1:26] All I do is deep brain stimulation [1:28] surgery and a complement to that is [1:30] focused ultrasound or transcranial [1:32] focused ultrasound, which is a [1:33] non-invasive way to do an ablation in [1:36] the brain, recently FDA-approved. And [1:38] it's FDA-approved for tremor at the [1:40] moment. Deep brain stimulation is a [1:42] procedure where we have to place a uh a [1:45] very thin wire that's insulated deep [1:48] into uh a part of the brain that's [1:50] involved in Parkinson's disease, for [1:53] example. Uh [1:55] but that's actually not the therapy. The [1:57] therapy is delivering electrical [1:59] stimulation through the tip of that wire [2:02] or one of the tips as there actually are [2:04] multiple contacts at the bottom of the [2:05] wire. They're very small. It's a bit [2:07] more like I have to implant a a tool to [2:10] to deliver you a medication, uh but that [2:13] medication is going to be in the form of [2:14] electricity and it's going to be [2:16] delivered into a very small region of [2:17] the brain. I'm very privileged to be [2:19] able to interact with the human brain in [2:21] this way. It's always in the [2:23] it with the goal of trying to provide [2:25] somebody with a meaningful therapy. But [2:27] when we deliver electrical stimulation, [2:29] these electrodes, while they might be [2:30] sitting in a very small region of the [2:32] brain, there are regions [2:34] within a few millimeters of where these [2:36] electrodes are that if stimulated could [2:38] cause a temporary very brief [2:40] side effect, a a moment of laughter like [2:42] you said, or a moment of panic. And of [2:44] course, we can just shut that electrode [2:46] off. [2:47] But often, these side effects [2:49] could be therapeutic. And actually, [2:50] that's how we have discovered ways to [2:52] use deep brain stimulation um not just [2:55] for movement disorders like Parkinson's [2:57] disease, but for example, patients with [2:58] Parkinson's disease that have a [3:02] psychiatric uh [3:03] uh [3:04] comorbidity like depression or [3:06] obsessive-compulsive disorder, a lot of [3:08] these patients are highly compulsive uh [3:11] and impulsive. Um [3:13] Sometimes, these problems actually melt [3:16] away. And we're trying to help their [3:17] tremor, but the patients also tell us [3:19] that their gambling issue has gotten [3:21] better or their mood has improved. And [3:23] why is that? Well, [3:25] you know, there's probably more than one [3:26] reason. You know, you can help [3:27] somebody's mood by making their tremor [3:29] go away, of course. But we see laughter [3:32] in the clinic sometimes. And And why is [3:33] that? And that's because we're [3:35] stimulating parts of the brain that are [3:37] not just involved in these motor [3:38] circuits, but they're also involved in [3:41] what we call a limbic circuit or or part [3:42] of the brain involved in emotion. And if [3:44] we learn how to [3:46] modulate those areas therapeutically [3:49] step by step, we can actually develop [3:50] these therapies for other indications [3:53] like depression. I would say the most [3:55] impressive and consistent effect we have [3:57] when we have a patient with tremor who [3:59] has been trembling for the past 20 [4:01] years, if we can deliver stimulation [4:02] through that electrode in the clinic, we [4:04] have immediate relief of tremor. And [4:07] that is the effect that inspired me to [4:09] be a neurosurgeon when I was in college. [4:11] I've never really wanted to do anything [4:12] else except help develop that type of [4:16] therapeutic for an other another kind of [4:18] symptom. [4:19] >> I'd love to learn more from you about [4:22] OCD. Could you perhaps just tell us what [4:24] is OCD? [4:25] >> Sure. [4:26] >> What are some brain areas involved? What [4:29] are the current range of treatments? And [4:32] what's the difference between someone [4:34] who is obsessive and somebody who has [4:36] true OCD? [4:37] >> My perspective on OCD may be a little [4:40] bit different than a psychiatrist who [4:42] who lives and breathes OCD and sees [4:44] patients every single day with OCD. [4:47] Uh I probably take care of a three to [4:49] five patients a year with deep brain [4:51] stimulation for obsessive-compulsive [4:53] disorder. So I don't see these patients [4:55] as routinely, but my laboratory is [4:57] geared as a researcher. Uh I'm very [5:00] focused on trying to [5:02] improve outcomes of deep brain [5:03] stimulation for for OCD. So I I do feel [5:06] I have expertise and and a perspective [5:08] to share. I do feel that as a [5:11] neurosurgeon, I am [5:13] obligated to better understand [5:15] where the obsessions in the brain come [5:17] from and how we can interrupt them to [5:19] stop the compulsion that's associated [5:21] with the obsession better than we're [5:23] actually doing it. I've been leading an [5:25] endeavor with a number of collaborators [5:27] around the country to try to [5:30] better understand the circuits in the [5:32] brain, uh study them in humans both [5:35] invasively and non-invasively. That [5:37] would be with an electrode-based [5:39] surgery, uh sort of like we do in [5:40] epilepsy to understand where seizures [5:42] come from. We want to understand better [5:43] where obsessions come from, but we're [5:45] also working with imaging experts and [5:47] geneticists to understand OCD [5:49] at a broader level as well. I consider [5:52] OCD to be a a spectrum disorder in a [5:55] way. And I [5:57] I I apologize to those who who might [6:00] feel that I'm using that term [6:01] incorrectly. I'm using it [6:03] in a way to describe [6:05] patients that have obsessions and even [6:07] some related compulsions might not meet [6:09] criteria for OCD. As a neurosurgeon, I'm [6:12] really obsessive about safety and [6:15] compulsive about my surgical procedures. [6:17] So, you know, I I think that some aspect [6:20] of OCD, which we often joke about, but [6:22] we should, you know, consider seriously [6:23] cuz people do suffer from this. [6:25] Uh some aspect of it helps us. [6:28] Uh there are, you know, famous uh CEOs [6:31] that probably have some level of OCD. [6:35] Uh surgeons and scientists alike. So, uh [6:38] perhaps if it can be controlled, it's an [6:40] asset. And uh [6:43] but if it goes awry and it's [6:45] uncontrollable, then it becomes [6:47] obsessive-compulsive disorder. And [6:50] uh I tend to see the patients that are [6:51] the most severe. So, they have failed [6:53] medication. And there are multiple [6:55] medications that are worth trying for [6:56] OCD. Some can actually be very helpful. [6:58] >> Which which neurotransmitter systems do [7:00] they tend to poke at? [7:02] >> Well, SSRIs are sort of the the first uh [7:05] line for OCD, but also tricyclics can be [7:07] helpful. So, this is still the serotonin [7:09] system. [7:10] Uh [7:10] but as we know, the serotonin system [7:12] interacts with the, you know, [7:13] noradrenergic uh system and the dopamine [7:16] system. So, it's hard to um [7:19] be specific to one of these things. And [7:21] I think that's also why it's hard for us [7:23] to predict how these medications are [7:25] going to to work for these kinds of [7:27] patients. But tricyclics and SSRIs can [7:30] be very helpful and are definitely first [7:32] line. And there's others. Exposure [7:34] response prevention is probably the most [7:37] effective option, which is kind of like [7:39] cognitive behavioral therapy, but these [7:40] are different and offered by [7:42] psychologists, and this is a whole [7:44] field, and there's a whole clinic at my [7:46] institution [7:48] uh focused was started by Edna Foa um [7:51] uh [7:52] at Penn who [7:54] this is what they do for these patients [7:56] uh is offer these types of cognitive [7:58] therapies, exposure to the stressor, and [8:02] to try to get patients to habituate to [8:04] whatever it is that stresses them and [8:05] causes these [8:07] uh compulsions to help these patients [8:09] live in everyday and function. The These [8:12] are all [8:13] fabulously helpful uh therapies for a [8:15] variety of patients, but there's still [8:17] about 30% of patients that still suffer [8:20] from OCD, and some of them have severe [8:22] OCD. Sometimes it's moderate to severe, [8:24] and those are the patients that I'm [8:25] really motivated to try to help. Um [8:28] our therapies for those patients right [8:30] now uh I would say are are worth [8:33] pursuing, but not optimal. Um [8:35] and so it's it's one of those things [8:37] that we have to balance as a researcher [8:38] because when you see patients like this, [8:41] you want to do everything you can to [8:41] help them, and I think it's important to [8:43] educate patients on the risk and [8:45] benefits of them. This is deep brain [8:47] stimulation surgery, but also [8:48] capsulotomy, which is more of an [8:49] ablation approach, a little bit like [8:52] deep brain stimulation, but rather than [8:53] delivering stimulation through an [8:55] electrode, you can actually heat the [8:56] tissue and even destroy it. Some would [8:59] say this part of the brain is very safe [9:00] to destroy. It's kind of like an [9:02] appendix. Um others would say it's safer [9:05] to modulate. I have seen uh patients do [9:07] very well with these ablations, and so [9:10] you know, you asked me earlier what what [9:12] I find so amazing about the brain, these [9:14] effects that we can have. Sometimes the [9:15] lack of effect is what's so amazing. You [9:18] can actually um [9:20] traverse parts of the brain without [9:21] having any adverse effects on patients' [9:23] um function at least that you can test. [9:26] Um but you can also destroy small parts [9:28] of the brain. We're talking 3 or 4 mm in [9:30] size. These little ablations can be [9:32] really helpful for patients, but have no [9:34] obvious side effects that we can tell, [9:36] perhaps after a short recovery from [9:38] surgery. Uh but nonetheless, despite how [9:40] safe they might be, uh these surgical [9:42] procedures still are surgical procedures [9:44] and patients are hesitant to proceed, [9:47] especially when they know that their [9:48] chance of a transformative effect is [9:50] quite low. We we can generally [9:53] um [9:54] uh [9:54] achieve a responder rate of about 50%. [9:57] Um and responders still have symptomatic [10:00] OCD. So, I'm really uh [10:03] uh sort of inspired to uh really find a [10:06] way to deliver these therapies in a more [10:08] disease-specific or symptom-specific [10:10] way. Were one to [10:13] come into your clinic this, you know, [10:15] for this sort of a work of ablations or [10:18] stimulation, [10:19] uh where would you first start to probe [10:21] in the brain? [10:22] Yeah, you this is a [10:25] uh [10:26] a disorder of both cortex and the sub- [10:29] subcortex. We find that areas in the [10:31] cortex like the prefrontal and [10:33] orbitalfrontal cortex are are not [10:35] functioning the way they would in a [10:36] non-OCD patient. They are often [10:38] hyperfunctioning, and we need to find a [10:40] way to try to normalize their function. [10:42] And then there are projections to the [10:43] subcortex. This is the basal ganglia, [10:45] the caudate, putamen, or the dorsal [10:47] striatum. And these are interconnected [10:48] with the ventral striatum. This is an [10:49] area of the brain that uh focus a lot of [10:52] my energy in. Um this is the ventral [10:54] striatum, which is not limited to but [10:56] includes the nucleus accumbens. Um this [10:58] is an area of the brain that [11:00] uh we know to be involved in gating [11:02] reward-seeking behavior. When it's [11:04] perturbed, it seems to gate compulsive [11:06] behavior, meaning a rat will pursue a [11:08] reward despite punishment, despite a [11:10] foot shock, for example. And that can be [11:12] similar to an OCD patient. They will [11:14] check their home for safety until 3:00 [11:17] a.m. in the morning and not sleep that [11:18] night. Doing something because of the [11:20] urge, but despite the risk. When our [11:23] judgment is consistently [11:25] uh sort of puts us at risk, that's where [11:28] we have something like OCD. [11:30] Contamination behavior, where they if [11:32] they feel contaminated they'll wash [11:33] their hands for hours repeatedly or if [11:35] they drop their toothbrush on the floor [11:37] this will lead to a compulsive behavior [11:39] of cleaning a toothbrush or brushing [11:41] your teeth consistently very very common [11:43] symptoms that we see or signs that the [11:45] patients report to us or or that we [11:47] observe. But you know patients with [11:49] eating disorders you know they tend to [11:52] if if they have bingeing disorder [11:53] they'll overeat if they have bulimia [11:55] they might purge despite the risk of [11:57] these things. And so [12:00] addiction is is similar we we tend to [12:02] drug seek if we're addicted [12:04] we'll we'll pay off a dealer in order to [12:07] get our fix and despite the risk and and [12:10] that type of urge despite the risk is [12:12] something that I I've always been really [12:14] interested in and it's a common [12:15] denominator to all of these problems and [12:18] if you think about these problems I mean [12:19] these are some of the most common [12:21] conditions in our society today and I [12:22] think the nucleus accumbens and the [12:24] cortical areas that we've been [12:25] discussing that that sort of send [12:27] projections to these areas are are [12:30] probably at least one of the main [12:32] circuits involved in these kinds of [12:33] things. [12:34] >> What is nucleus accumbens? What roles [12:36] does it play in healthy brain behavior [12:38] and in pathology? [12:39] >> Yeah the nucleus accumbens is a part of [12:42] the brain part of our reward circuits. [12:45] It has a lot of functions. Uh [12:48] it interconnects with many parts of the [12:50] brain. So when I started getting [12:52] interested in [12:54] reward and what a what I could do as a [12:56] surgeon to try to improve how we manage [13:00] rewards. [13:01] And what I mean by that specifically is [13:03] if you have an urge for a reward [13:06] that that's a normal phenomenon. That [13:08] that's not something we're trying to [13:09] stop. The the issue is if you have an [13:11] urge for a a reward that either puts you [13:13] or somebody else at risk it's probably a [13:16] reward we shouldn't have. If you're a [13:17] drug addict and you use heroin or opiate [13:22] that opiate might make you feel better [13:24] cuz life is stressful but the risk of [13:26] doing those things is really high. In [13:29] fact, potentially lethal. If you have [13:30] OCD and you [13:34] can't sleep at night because you're so [13:36] nervous that you didn't lock the door [13:37] and you've checked 30 times. That's an [13:39] urge we got to treat. Eating disorders [13:41] the same. This problem can be [13:43] ameliorated or improved upon by a better [13:46] understanding and a tailored treatment [13:49] to the nucleus accumbens specifically. [13:51] It seems that repeated exposure to [13:53] something like a drug of abuse or any [13:55] type of reward that is a really strong [13:59] reward. [14:00] In a way, it it can hijack normal [14:02] functioning of the nucleus accumbens. [14:04] So, the goal is to just disrupt [14:06] perhaps what is kind of habitual. [14:09] Um [14:10] or or at least this [14:12] kind of recurring problem that is [14:14] happening. You know, people that have [14:15] binge eating disorder, at least at a [14:17] severe level, they tend to binge about [14:18] once a day. [14:19] So, what we decided to do in the [14:22] operating room was [14:23] to actually try to leverage a tool that [14:27] we use all the time when we take care of [14:28] patients with Parkinson's. [14:30] So, with Parkinson's, these a lot of [14:31] these patients, not all, have tremor. [14:33] And so, when we place an electrode into [14:36] this motor structure [14:38] to try to improve their movement [14:39] disorder, [14:40] uh we often can hear tremor cells. [14:43] And they sound we convert their [14:46] electrical signal to an audible signal [14:47] so we can actually hear it. And it [14:49] sounds kind of like the tremor looks, [14:50] like the frequency of the signal is the [14:52] same as the hand shaking. [14:55] Exactly. [14:56] >> And you're poking around in a dedicated, [14:58] careful way, of course. [14:59] >> Yes. One poke at a time. [15:00] >> One poke at a time and with a very fine [15:02] wire set of wires, listening to the [15:04] electrical activity until you [15:07] you encounter some cells that are [15:10] sending out electrical activity at the [15:12] similar frequency. [15:13] >> Exactly. [15:13] >> And then you can stimulate them or quiet [15:16] them and see if the tremor goes away. [15:17] >> So, we we are very confident that when [15:19] we stimulate that area of, in this case, [15:22] the subthalamic nucleus, we will disrupt [15:24] that tremor circuit and that tremor will [15:27] dissolve. And it does. [15:28] >> So, what is the analog to tremor in [15:32] terms of appetite and desire to binge? [15:34] >> Craving. [15:35] So, craving is a term [15:38] that, you know, there's probably other [15:39] terms we could use, by the way, but that [15:41] that's the term we've chosen to use for [15:42] a number of reasons. One, because people [15:44] relate with that term. People that have [15:46] binge eating disorder or obesity, they [15:49] if you ask them if they crave, the [15:51] answer will often be yes. Um [15:53] if you ask them if they lose control or [15:54] binge, they might not know what you [15:56] mean, or they might not actually feel [15:58] out of control even when they are. [16:00] Um [16:01] So, uh [16:02] but the word craving is relatable. And [16:05] so, we set out to see if we could [16:07] identify craving cells. [16:09] Um in a patient with OCD, which is [16:13] related, in fact, we target a very [16:14] similar part of the brain, [16:16] uh [16:17] we tried to identify [16:20] cells related to obsessions. And we [16:22] believe we did do that. It was a single [16:24] case study uh where we tried to optimize [16:26] where our electrode was placed. So, we [16:28] had some [16:29] proof of concept that we would be able [16:31] to elicit a sort of disease-specific [16:34] symptom in the operating room, assuming [16:36] the patient could tolerate being awake. [16:38] Not everybody needs to be awake for this [16:40] procedure, but at least for these first [16:41] in human trials where [16:43] um we're trying we're trying to [16:45] establish where in the brain we need to [16:46] be, uh I think this type of approach is [16:49] really critical. [16:50] >> What is the status of non-invasive brain [16:52] stimulation, ablation, and blocking [16:54] activity in the brain? My understanding [16:56] is that transcranial magnetic [16:57] stimulation is being used to treat [16:59] depression and a number of other [17:01] um brain syndromes [17:03] uh non-invasively, so no no drilling [17:04] through the skull. My understanding is [17:06] that the spatial precision isn't that [17:08] great. [17:08] >> Yes. [17:09] >> Um [17:09] ultrasound is something I hear a lot [17:12] about these days, um and my [17:14] understanding is that ultrasound can [17:15] allow researchers and clinicians to [17:17] stimulate specific brain areas. What are [17:19] your thoughts on these forms of [17:21] non-invasive, meaning no no flipping [17:24] open of a piece of the skull, type brain [17:26] stimulation and blockade of brain [17:29] activity? [17:30] >> We need to embrace non-invasive [17:31] approaches. Some of them are a little [17:33] fluffy in that we don't understand how [17:34] they work. [17:36] We don't necessarily understand how deep [17:37] brain stimulation works, by the way, so [17:39] but because we don't know exactly how [17:40] they work, they're not as precise as we [17:42] would like them to be, so we have work [17:43] to do there. And I actually think that [17:45] work is doable and actually underway. [17:47] TMS, transcranial magnetic stimulation, [17:50] it is FDA approved for depression, by [17:52] the way. It's also FDA approved for OCD [17:54] and for nicotine addiction. We believe [17:55] we can use TMS to to define a circuit [17:58] that if modulated improves OCD, albeit [18:03] temporarily. And in those patients, if [18:05] it's temporary, they would be [18:06] appropriate for an invasive study. So, [18:09] um something we're actively working on. [18:11] I've always believed that neurosurgeons [18:12] need to be part of the discussion with [18:15] these non-invasive approaches. We don't [18:16] need to do them, um [18:18] but um I think we can help make them [18:21] more precise and to probe non-invasively [18:24] with purpose. Perhaps one day there will [18:26] be a TMS target for anorexia and [18:29] obesity. [18:30] Uh [18:31] if we are scratching the surface with [18:34] invasive approaches to these problems, [18:36] we we're even doing less with the brain [18:39] stimulation. Um so, we have so much work [18:41] to do there. Eating disorders and TMS [18:44] have been so [18:45] um sort of [18:47] uh scarcely studied or or [18:49] there have been such little research [18:50] done in that space. Um so, it it it is [18:52] an area that we need to to work on. So, [18:54] ultrasound right now [18:56] transcranial magnetic guide [18:59] magnetic resonance guided focus [19:01] ultrasound. So, um [19:03] uh this this is an FDA approved method [19:07] to [19:09] deliver an ablation to the brain [19:11] non-invasively. [19:13] There are uh researchers, myself [19:15] included, that are trying to use [19:17] transcranial magnetic guided magnetic [19:19] resonance guided focused ultrasound or [19:20] MRI guided focused ultrasound [19:23] to use it in a modulatory way, not just [19:25] as an ablation, but to drive neuronal [19:28] activity or inhibit it, perhaps. We're [19:30] still learning how to do that. Um [19:32] there are trials that are trying to [19:34] understand if you can use ultrasound to [19:35] open the blood-brain barrier so you can [19:37] deliver a medication to that specific [19:40] area. Perhaps for a brain tumor or [19:42] something like that. [19:44] So, it's a very exciting field [19:47] and it is FDA approved for tremor right [19:49] now. And so, I actually [19:51] do it routinely um [19:53] for patients with tremor with [19:54] Parkinson's or essential tremor. [19:56] And so, I I love doing it. It's often [20:00] just kind of a miracle because there's [20:02] no incision. I don't have to place an [20:03] electrode into the brain to achieve a [20:05] similar result. It's fabulously [20:07] effective for these patients. It treats [20:08] patients on one side, usually their [20:10] dominant hand or their worse hand. And [20:12] it um [20:13] it really speaks to the fact that well, [20:15] you can deliver non-invasively an [20:17] ablation to the brain in a hypothesized [20:20] zone that we think is related to the [20:22] problem at hand. And at least with [20:24] tremor, it works really well. Could this [20:26] be effective for psychiatric disease, [20:29] obesity, eating disorders? Uh [20:32] well, [20:33] perhaps. Uh actually, that would be the [20:35] ideal. [20:36] The problem is we don't know where to do [20:38] the ablation. [20:39] Um there is a trial that we would like [20:42] to do for OCD where we would deliver an [20:44] ablation to the same area of the brain [20:46] that we've been delivering ablations to [20:47] for years for patients with OCD and it [20:49] helps a bit. That's called a [20:50] capsulotomy. Um [20:52] but really the outcome is probably going [20:54] to be about the same. It's a nice method [20:55] because it's it's not invasive, but we [20:57] need to find a new target for these for [20:59] these conditions. And because of the [21:01] common denominator of the urge despite [21:05] the risk, sort of that compulsion, [21:07] Um yeah, perhaps it could be the same [21:09] target. I don't know. Um [21:11] but I would argue we need to do these [21:13] modulatory experiments either with a [21:15] device or with invasive recordings uh to [21:19] better understand where these problems [21:21] are coming from, to define where we [21:23] should do an ultrasound treatment. There [21:24] has been a revolution in America. It was [21:27] in Europe before it was in America where [21:29] we would do stereo encephalography, [21:31] which is basically like doing an EEG of [21:34] patients with epilepsy, but with [21:36] invasive electrodes. And we would place [21:38] tiny little wires less than a millimeter [21:40] diameter all throughout the brain into [21:42] parts of the brain that we believe are [21:43] involved in seizures. [21:45] And we would admit the patients to the [21:46] hospital and figure out where the [21:48] seizures were starting and propagating. [21:50] And then um yeah, we could stimulate [21:52] through these electrodes to see if there [21:53] was a symptom that was important and I [21:56] try to identify a region that we thought [21:58] we could either [21:59] remove surgically, ablate with a laser, [22:02] or put a stimulator in it, perhaps. [22:04] Um that's commonplace now for epilepsy. [22:07] Um [22:08] and it works extremely well and it's [22:10] very safe. Of course, it's still a brain [22:12] procedure, um but the uh the [22:14] complication rate is surprisingly low, [22:17] quite honestly, for the amount of [22:18] electrodes that we place. And it's [22:20] extremely well tolerated. Most of these [22:21] patients leave the hospital and they [22:22] don't even feel like they've had [22:23] surgery. [22:25] So, uh there's actually a lot of [22:26] interest in using that procedure to [22:28] study [22:30] mental health disorders. We are trying [22:31] to do it for patients with obsessive [22:33] compul- compulsive disorder. We're [22:34] awaiting an FDA decision on that. Uh but [22:37] actually I credit uh [22:39] our colleagues at Baylor and at UCSF for [22:42] for studying this uh already. Bringing [22:44] together the epilepsy technique and the [22:46] psychiatry expertise to study how we [22:48] could better target electrodes in [22:51] depression. [22:52] And I'll tell you, if they have a [22:53] consistent target, [22:55] perhaps there becomes an ultrasound [22:58] target. Um but right now the approach is [23:01] a bit more reversible cuz you can always [23:03] shut that electrode off or even remove [23:05] the electrode if perhaps it's not in the [23:08] optimal location to treat the [23:09] depression. [23:10] Uh [23:11] but actually after a large volume of uh [23:13] cases perhaps, they could pool that data [23:17] to develop a a new ultrasound target for [23:20] depression. I think that would be [23:21] fabulous and probably is their long-term [23:23] goal. Not to speak for them, uh but that [23:25] would be something that I I'm sure is on [23:26] their radar. [23:28] You might ask, well, why aren't you [23:29] doing this for obesity right now in uh [23:31] in our in our study? And the reason is [23:33] that um [23:34] we've developed a target for obesity uh [23:36] and binge eating disorder uh developed [23:38] out of mice that we believe um is [23:41] relevant for the human state because you [23:43] can model this problem in a mouse a bit [23:46] better than you than you can model [23:47] depression or OCD. [23:49] So, we feel like we can rely on the [23:51] preclinical studies more. [23:53] Whereas with these perhaps more I don't [23:55] want to say more complicated, but more [23:57] human mental health conditions that are [23:59] hard to model in a mouse, you really [24:02] have to study it in the human. And you [24:05] can perhaps start in an epileptic [24:06] patient, a patient that has electrodes [24:08] and try to provoke a depressed state or [24:10] study epileptics that have comorbid [24:12] depression, for example. [24:14] Uh and that can really validate this [24:15] approach as well. But in the end, it's [24:17] it's getting into the human brain that [24:19] we need to do in the disease [24:21] specifically. [24:23] That will eventually lead to a [24:25] non-invasive approach, either a lesion [24:28] or modulatory approach. Modulatory would [24:30] be like TMS or lesion approach would be [24:32] with uh ultrasound. [24:33] >> If people can be made to feel or make [24:36] themselves feel just a little bit [24:38] better, [24:39] a little less anxious just prior to a [24:41] craving episode or a binge episode, [24:44] maybe even if people can become better [24:46] at detecting their own internal states [24:49] and when they're kind of veering toward [24:50] a binge or veering toward using a drug [24:53] or [24:54] maybe even veering toward suicidal [24:57] thinking. Seems like that awareness [24:59] seems like maybe among the best tools [25:02] that people could develop. [25:03] >> Yes. [25:04] I've always thought that if we can [25:05] improve awareness, we can improve [25:07] outcomes. I think that's probably true [25:09] for many of these patients. The problem [25:11] I think comes down to the fact that some [25:13] of these patients are so resistant to [25:15] treatment. [25:17] And the patients that we see as a [25:19] surgeon, for example, are the patients [25:21] that they've tried cognitive behavioral [25:24] therapy, and they've tried medications, [25:26] they've tried behavioral management. [25:28] They're as aware as they could possibly [25:29] be, and they still lose control. We've [25:32] had this studied in the lab. So, we will [25:35] bring patients to the laboratory with [25:36] this implanted device [25:38] to to try to provoke this electrographic [25:41] electrical signal [25:42] um that can be detected by the actual [25:44] device that will stimulate them when [25:47] they're at home. [25:48] But before we actually initiate [25:49] stimulation, we want to to see can this [25:51] device detect this craving cell signal, [25:56] which is going to be different than what [25:56] we saw in the operating room, because [25:57] that's a single cell, but these devices, [26:00] these electrodes are about a millimeter [26:01] in diameter instead of like a tenth of a [26:03] millimeter, which is what we use in the [26:04] operating room. Um so, they're they're [26:07] only hearing or or detecting, I should [26:10] say, thousands of cells' responses. [26:13] And [26:15] we actually have a way to provoke [26:18] binges. It's called a mood provocation. [26:20] It's very well well very well validated. [26:22] It's a little bit like provoking [26:23] seizures in the epilepsy monitoring but [26:26] here in these sort of [26:27] uh [26:28] psychiatric monitoring unit or the the [26:30] food monitoring unit, [26:32] uh we we actually have a psychiatrist [26:35] and eating disorder specialist come and [26:37] induce a mood that is related to [26:41] each patient's sort of self-described [26:44] binge episode. [26:46] >> So, the psychiatrist comes in and [26:47] provokes [26:48] >> Yes. [26:49] >> a feeling that can evoke the negative [26:53] behavior. [26:54] >> That's exactly right. So that we can [26:55] video and synchronize the video to the [26:58] brain signal recordings. [27:01] The patients all wear an eye tracker so [27:02] we can see what they're eating at all [27:04] times and what they're looking at [27:05] specifically and that allows us to [27:10] have the best temporal resolution [27:12] possible to understand what is happening [27:14] right before the bite. [27:16] And even under video surveillance [27:18] through a one-way one-way mirror in a [27:20] laboratory setting when patients are [27:24] very well aware that they're [27:26] there to be studied if they're going to [27:28] binge. [27:29] They still do [27:31] and we believe they do because they just [27:33] can't control it as aware as they are of [27:36] it. [27:37] And it's probably because they're the [27:38] most severe. [27:39] So I think if we can improve awareness, [27:42] not just the societal awareness that [27:43] I've been talking about earlier, but the [27:45] patient awareness [27:46] around their problem, I think that could [27:48] be a powerful way to help so many of [27:50] these patients and that's sort of the [27:51] role of cognitive behavioral therapy. [27:54] The problem with cognitive behavioral [27:56] therapy or I should say the limitation [27:57] of it, I actually don't have a any [27:58] problem with it. I think it's a [27:59] wonderful treatment. [28:03] Is that if you stop it, many of these [28:04] patients go back to their old behaviors. [28:07] I don't want to say old habits, but it [28:08] might be a habit, but the old behaviors. [28:11] And so um [28:13] that's the problem is it's not [28:14] necessarily lasting in the absence of [28:15] continued cognitive behavioral therapy. [28:17] Some people can benefit [clears throat] [28:18] from it long term, but some can't. Uh [28:21] but I think in in in in the less severe [28:23] patients, [28:25] improving awareness key. [28:26] But in these really refractory patients, [28:28] this is this is kind of like this is the [28:29] disease despite the awareness. They [28:32] can't control themselves and that's what [28:34] we're trying to restore is that improved [28:36] ability to control their behavior. [28:38] >> Do you think there's a role for machines [28:40] and artificial intelligence here? [28:44] There are a couple laboratories up at [28:45] the University of Washington that are [28:47] using [28:49] particular signature patterns of within [28:52] voice [28:54] to try and help [28:55] suicidal [28:57] people who are suicidally depressed [28:59] know when they're headed towards an [29:01] episode before they even can consciously [29:03] know. This gets right down to issues of [29:04] free will and whether or not machines [29:05] can be smarter than we are. But, you [29:07] know, one could argue that some of the [29:08] search algorithms [29:10] on Google and other search engines are [29:11] actually more aware of our preferences [29:14] than we are. [29:14] >> Mhm. [29:15] >> Um [clears throat] [29:16] Basically, what these are these are [29:17] devices that are listening to people [29:19] talk all day. They're also paying [29:20] attention to patterns of breathing and [29:22] how well people slept, etc. Integrating [29:24] a a huge number of cues and then [29:26] signaling somebody with a uh you know, a [29:28] yellow light. Like, you know, you're [29:30] headed into a depressive episode. The [29:32] person might say, "Oh, I feel fine." Or, [29:33] "I feel pretty good. This is kind of [29:35] baseline state for me." And they're [29:36] saying, "Uh-uh. This is where you were [29:40] preceding the last episode that took you [29:42] down a deep dark trench and it took [29:43] months to get out of." [29:44] >> Yep. [29:45] >> Um I wonder whether or not some of these [29:47] devices could help with the sorts of [29:49] things that we're talking about today. [29:51] >> Yeah. [29:51] I think so. Um [29:53] I've always said we have to get in the [29:55] brain before we get out of it. And if we [29:57] get in the brain and understand what [29:58] these signals look like, we'll know what [30:01] those non-invasive signals are. [30:03] I think it's possible that we are uh [30:07] scientifically sophisticated enough to [30:10] use machine learning and sort of this [30:12] kind of bot technique to anticipate when [30:16] somebody is going to be highly [30:18] impulsive. You know, suicide is the most [30:19] dangerous impulse. It's something that [30:21] is [30:23] immensely a focus of the lab is [30:26] impulsivity. We've talked mostly about [30:27] compulsion. [30:29] Compulsion being, you know, going after [30:31] a reward or or the urge despite the [30:33] risk. Um [30:35] >> [clears throat] [30:35] >> Impulsivity is is similar, but [30:37] different. It's It's kind of going after [30:38] something um [30:41] a little bit if you if you model [30:42] impulsivity in a in a mouse, it's, you [30:44] know, related to [30:46] you know, going after a food reward [30:49] without the sort of paired tone that [30:52] you're the mouse is supposed to wait [30:53] for. The mouse doesn't want to wait [30:55] anymore. They they just go after the [30:56] food. Um [30:57] >> I've been that mouse. [30:58] >> Yeah, we've all been. [31:00] We can all relate with this uh to a [31:02] certain extent. Again, it's a spectrum. [31:04] >> [laughter] [31:05] >> So, [31:06] um [31:07] So, in any case, I [31:08] non sequitur, but I I I certainly think [31:11] that there is a way to use our own [31:14] body's physiology [31:16] to anticipate [31:19] when these impulses are coming online. [31:21] How best to do that, I think we're just [31:23] scratching the surface, but um [31:26] these are the kinds of solutions we [31:27] need. Some of these problems are of [31:29] epidemic proportions, largest public [31:31] health problems in this country, in this [31:33] world, obesity, opiate crisis, [31:35] depression, suicidality. I mean, that's [31:38] like a third of our country, maybe more. [31:40] We need scalable solutions. But, you [31:41] know, I'm I'm a neurosurgeon. I'm only [31:43] going to be able to treat the most [31:45] severe of patients with these problems. [31:47] You know, how [31:48] you know, we've only done about 200,000 [31:51] deep brain stimulation surgeries [31:53] ever. [31:54] So, I mean, the problem we're talking [31:56] about here is [31:58] 50 million Americans. There's no [32:00] possibility that surgeons can address [32:02] that problem. But, we could help inspire [32:06] an initiative to go after that kind of [32:08] problem or help make it more rigorous. [32:10] Because the last thing we need is a you [32:13] know, some sort of wearable fancy tool [32:14] that you know, [32:18] waste people's money and time. You know, [32:19] we need real therapies for these things. [32:21] Not that these devices that we're [32:23] discussing are not. I think actually [32:24] there's lots of promise. [32:26] We use machine learning in the lab all [32:27] the time. I'm not a [32:29] an electrical engineer or the [32:31] computational neuroscientist doing this [32:32] type of work. I I just helped develop [32:34] the hypotheses around it. But, um I help [32:36] fundraise around it. But I I definitely [32:39] think there's a future for it. I just [32:41] suspect we're we're scratching the [32:42] surface on how best to do it. [32:44] >> I really appreciate you sharing those [32:46] tools. A number of people [32:48] I'm guessing out there might want to [32:49] become neurosurgeons. I really believe [32:51] that in hearing today's conversation [32:53] that you will spark an interest in [32:55] medicine and or neurosurgery. [32:57] >> so. [32:57] >> Um well, certainly you need to be a [32:58] physician before you can become a [32:59] neurosurgeon. So and neurosurgery in [33:01] some cases and that would be beautiful [33:03] and I predict that will be happen that [33:05] will happen excuse me as a consequence [33:07] of what you've shared today. Really want [33:09] to thank you for taking time out out of [33:10] your not just immensely busy, but very [33:12] important schedule because [33:14] again, the work that you're doing is [33:16] really out there on that cutting I don't [33:19] want to say bleeding edge cuz in this [33:20] context it's not going to sound right, [33:21] but on that extreme cutting edge of what [33:24] we understand about how the human brain [33:26] works and how it can be repaired. So on [33:28] behalf of everybody and and myself as [33:30] well, thank you so so very much. [33:32] >> I'm honored. Thank you so much for [33:33] having me.