你的大脑与...脑机接口

这个设备开着吗？

Is this thing on?

这是一位名叫凯西·哈雷尔的男子在参与脑机接口试验时提出的问题。尽管ALS（肌萎缩侧索硬化症）让凯西失去了控制说话肌肉的能力，这项技术却让他重新获得了说话的能力。

That's what a man named Casey Harrell asked while participating in the trial of a brain computer interface which though ALS had caused Casey to lose control of the muscles used to speak allowed him to speak again.

那是一个无比震撼的时刻，我们非常激动且荣幸能与实现这一突破的科学家、神经外科医生和神经工程师们讨论这个项目。

It was an incredibly powerful moment and we're so excited and honored to be discussing the project with the scientists neurosurgeons and neural engineers who made it happen.

我们是艾莎医生和迪恩·谢尔扎伊医生，这里是《你的大脑与脑机接口》节目。

We're doctors Aisha and Dean Scherzai and this is your brain on brain computer interface.

迪恩，这期节目太令人兴奋了，特别是在当今时代，我们正见证着这一领域的飞速发展。

Dean this is such an exciting episode especially because of how at this day and age we're seeing the advances that are happening in this area.

作为一个《星际迷航》粉丝，我从小看着这部剧长大，是个十足的科幻迷，你也是。看到这些曾经只存在于故事和传说中的事物成为现实，并见证它对患者生活产生的改变，真是令人无比振奋，在很多方面都让人肃然起敬。

And as a Trekkie, I grew up watching Star Trek, and I'm a big sci fi fan, and you are as well, To see some of these things that were basically related to stories and myths coming to reality and seeing its impact on patients' lives is incredibly exciting and awe inspiring in many ways.

确实如此。

It is.

特别是作为神经科医生，我们经常在急诊室见到因事故导致瘫痪的患者——脊髓损伤或枕叶/眼部受伤使他们余生都无法行动或视物，或是中风后失去语言能力，又或是ALS患者逐渐丧失说话和行走能力。

Especially as neurologists we see patients who come to the emergency room who've had an accident and are paralyzed for the rest of their life because of a spinal cord injury or injury to their occipital lobe or eyes and they can't see or the fact that they had a stroke and they can't speak, or they have ALS and they slowly lose their ability to speak or walk.

我们正站在医学革命的风口浪尖上，以前不能行走的人现在可以走了，不是可能，他们必将能够行走。

We are at the precipice of a remarkable revolution in medicine where people couldn't walk before could potentially walk, not potentially, they will be walking.

是的，那些失去语言能力的人，我们现在看到的这个案例，非常令人振奋。

And those who had lost their capacity to speak, that one we're seeing now, it's very exciting.

本期节目的两位嘉宾是神经外科医生。

Two of our guests for this episode are neurosurgeons Doctor.

David Brandman和神经工程师。

David Brandman and neural engineer Doctor.

Sergei Staviski。

Sergei Staviski.

两位都是加州大学戴维斯分校神经假肢实验室的联合主任，该实验室开发了我们开场提到的技术。

Both are the co directors of the UC Davis neuroprosthetics lab, which developed the technology we mentioned in the intro.

以下是他们近期视频中的一个片段。

Here's a snippet from one of their recent videos.

这感觉非常像我。

It feels a lot like me.

这让许久未听到我声音的人们落泪。

It makes people cry who have not heard me in a while.

我认为这项技术已足够成熟，能真正让他融入社会并正常交流。

I think we're at a point where it does work well enough that it really lets him just be a part of society and to talk.

我仍能全职从事气候行动工作，借助这个设备，我可以高效出色地完成工作。

I am able to still work full time for climate action, and I am able to do that well and efficiently with this device.

我希望现在像我这样的人都能拥有这样的设备来辅助沟通，获得与我同等的机会。

I hope that we are at a time when everyone who is like me have the same opportunity as I do to have a device like this that will help them communicate.

让我们一起实现这个目标，好吗？

Let's all make that happen okay.

布兰德曼医生提到的一个有趣现象是，

One of the interesting things that Doctor.

社会对脑机接口和这类医疗技术的认知仍深受科幻作品的影响。

Brandman mentioned was how society's perception about brain computer interface and this technology in medicine is still influenced by their perception of what sci fi is.

我觉得很多事物都存在这种情况。

And I I think that's the case with a lot of things.

通常新概念最初都是以故事和神话的形式出现，然后不知不觉间就成为了我们现实的一部分。

Usually newer concepts are introduced as stories and as myths and then suddenly before you know it they become a part of our reality.

这确实会在人们心中引发某种忧虑和恐惧。

That does bring about some sense of apprehension and fear in people.

确实如此。

It does.

这很自然，尤其是在大众层面上，人们对新事物总会持怀疑态度。

That's natural especially at the population level for people to be incredulous and suspicious of new things.

这几乎像是一种防御机制。

It's almost like a defense mechanism.

从进化角度来说，我们更倾向于自我保护。

Evolutionarily we are more on the side of protecting ourselves.

这就是为什么每当新事物出现时总会遇到阻力。

That's why anytime something new comes up there's resistance to it.

抗生素问世时就曾遭到极大质疑。

When antibiotics came there was a huge suspicion.

X射线问世时，人们对其用途和实际效用充满疑虑。

When x rays came there was a suspicion of what it was being used for and whether it was useful.

其他成像技术和基因技术也是如此。

Same thing with other imaging techniques and genetics.

你会听到各种关于基因技术的传言，很多与现实毫无关联，但人们对基因检测这类技术仍相当怀疑。

You hear all these stories about genetics, really things that don't even have any connection to reality, but people are quite suspicious of genetic testing and things of that nature.

每个发展阶段都伴随着担忧和猜疑。

And every step has worries and suspicions.

部分原因在于我们过去在向大众传播科学方面存在不足。

And part of it is that we have failed in the past, really translating science to the population.

这正是问题所在。

That's where the failure has been.

我们从科学角度将自己与大众割裂开来，突然推出新技术时却缺乏必要的背景铺垫。

We separate ourselves from a scientific perspective, from the population, and all of a sudden something is introduced without the necessary storylines that have to be built around something.

而关于脑机接口的叙事，细想起来全是负面内容。

And the storylines around brain machine interface, if you think about it, it's all negative.

《星球大战》，博克。

Star Wars, Bork.

天啊。

Oh my gosh.

那个

The

半机械人？

cyborg?

半机械人。

The cyborg.

在电影里，通常都是被外部力量控制的那个反派角色。

In movies, it's usually that villain that is being controlled by an outside force.

光是想到你的大脑被你自己以外的其他东西控制，以及整个隐私问题——这些信息如何可能直接公之于众，这本身就非常可怕。

And just the thought of your brain being controlled by something else other than you and the whole privacy issue of how things can actually just become available to public, that's also very scary too.

你说得对。

And you're right.

我认为我们在这方面对公众的解释工作做得不够好。

I don't think we've done a good job of explaining this to the public.

这次对话的整个目的就是向大家——我们可爱的听众们——介绍这个概念和技术，让他们了解所有这些正在进行的精彩工作。

The whole purpose of this conversation is to introduce this concept and this technology to everyone, our lovely listeners, and for them to be aware of all this fantastic work that is being done.

是的。

Yes.

这可能听起来令人不适且新奇，大脑中植入东西的整个想法听起来很具有侵入性。

It may sound uncomfortable and new, and the whole idea of something inside your brain sounds intrusive.

但当你看到结果，看到它如何影响人们的生活时，我认为这部分可能是最具人性化的事情了。

But when you look at the outcome and when you look at how it impacts people's lives, I think that part of it is probably one of the most human thing ever.

我们正在提升人的尊严、尊重、生活质量，以及他们成为想成为之人的能力。

We're augmenting human dignity and respect and their quality of life and their capacity to be who they want to be.

绝对美妙。

Absolutely beautiful.

我觉得这极其美好且充满希望，对于我们的演讲者即将深入探讨这个话题，我感到非常兴奋。

I find that incredibly beautiful and hopeful, and I'm so excited for our speakers to actually dive into this.

虽然我们的工作重点一直放在预防上，但这项技术不仅能让我们实现预防，还能进行检测和早期发现。

Though much of our work has been focused on prevention, this technology gives us access not just to prevention, but also detection and early detection.

这至关重要，因为它能每分钟、每秒实时监测，因为它与你的身体相连。

And that's critical because it can do that minute by minute, second by second as it's connected to your body.

这有点吓人，但也令人难以置信地充满力量。

That's a little scary, but it's also incredibly empowering.

借助这些新工具，我们检测疾病的能力呈指数级提升。

Our capacity to detect disease moves exponentially with these new tools.

现在我们必须谨慎，但类似的担忧我们以前也有过。

Now we must be cautious, but we've had these fears before.

在19世纪，我们发明了听诊器。

In the eighteen hundred's, we found stethoscope.

人类首次能够听到他人的心肺声音。

For the first time, could hear people's hearts and lungs.

当时人们认为这是种魔法，是邪恶的，并为此感到担忧。

And it was thought that this was a magical thing and it was evil and people were worried about it.

在二十世纪初，我们发现了X光，能够真正透视人体内部，看到骨骼和组织。

And in the early nineteen hundred's, we discovered x-ray, where you could actually go inside of a person's body and see their bones and tissue.

到了七十年代，我们又有了CT和MRI技术。

And in nineteen seventy's, we detected CTs and MRIs.

现在我们可以看清人体内部的一切。

Now you could see everything inside somebody's body.

这本身就令人畏惧。

That's scary in itself.

接着我们迎来了基因组学和蛋白质组学，让我们能够真正了解构成我们的基本物质和编码方式。

And then we got genomics and proteomics where we can actually know the very fiber, the very substance of what we are made of and how we are coded.

这简直不可思议。

That's incredible.

这是下一步发展。

This is the next step.

一步。

Step.

如今我们有了可穿戴设备，能告诉你睡眠质量、深度睡眠时长、行走步数、呼吸模式，甚至压力模式。

And now we have wearables that can tell you how you slept, the depth of your sleep, how much you walked, your breathing patterns, even your stress patterns.

这些都是非凡的工具。

These are remarkable tools.

这款设备将技术提升到新高度，因为它能直接与你的身体连接。

This one takes it to the next level because it connects directly to your body.

对于那些功能缺失的人，它能帮助他们恢复功能。

And for those that don't have a function, it gives them that function back.

疾病的早期检测将变得无处不在。

Early detection of disease becomes ubiquitous.

它将始终存在，这非常强大。

It becomes ever present and that's powerful.

为了进一步探讨脑机接口的过去与未来，我们很荣幸邀请到首位嘉宾神经学家。

To talk more about the past and future of brain computer interface, we're excited to welcome the first guest neurologist, Doctor.

李·霍克伯格博士。

Lee Hochberg.

医生。

Doctor.

霍赫伯格是麻省总医院神经技术与神经康复中心的主任，这听起来非常棒，同时也是BrainGate神经接口系统试点临床试验的首席临床研究员。

Hochberg is the director of the Center for Neurotechnology and Neurorecovery that sounds awesome at Massachusetts General Hospital and principal clinical investigator of the pilot clinical trial of the BrainGate neural interface system.

医生。

Doctor.

霍赫维克花费了超过二十年时间，致力于为严重瘫痪患者（如ALS或脊髓损伤患者）恢复沟通和行动能力的研究。

Hogwerk has spent over two decades working on ways to restore communication and mobility for individuals with severe paralysis such as those affected by ALS or spinal cord injuries.

他的见解不仅涉及脑机接口背后的技术。

His insights don't just address the technology behind brain computer interface.

他还以对临床实施和人类自主权的深刻尊重来探讨伦理问题。

He also discusses the ethics with a profound respect for clinical implementation and human autonomy.

以下是他的谈话内容。

Here's his conversation.

我对神经科学和医学的兴趣始于很久以前，但对脑机接口这一领域产生兴趣，是在我申请医学博士项目时，当我从布朗大学毕业时，遇到了一些思考该领域现状的杰出人物。

My interest in neuroscience and medicine started long ago, but the interest in what developed to be this field of brain computer interfaces was really, as I was looking around at MD PhD programs, as I was graduating from Brown got to me some incredible people who were thinking about what the state of the field was at that point.

许多人没有意识到的是，美国国立卫生研究院自60年代末70年代初就设立了神经控制实验室和神经假体项目。

And what many people don't realize is that the NIH has had a laboratory of neural control and a neural prosthesis program dating back to the late 60s and early 70s.

早在70年代就有一些重要的早期论文和会议，当时人们就公开预言未来瘫痪患者可能重获肢体控制能力——因为我们能从大脑记录运动意图信号，例如跨越颈髓损伤的断连处重新连接这些信号，进而刺激肢体恢复运动功能。

And there's some great early papers and some really important conferences that date back certainly to the 70s, where people were saying out loud about our future where somebody with paralysis might regain the ability to control their limbs, because we'd be able to record their intent to move from the brain and reconnect those signals, for example, across a disconnection across a cervical spinal cord injury, and then stimulate limb and re enable movement.

因此在这种背景下，当我正在规划自己的专业发展方向时，埃默里大学有位生理学教授表示，他确信并成功获得了NIH首批专注于神经假体项目的合约之一。

So in that context, as I was trying to figure out where my training was going to continue, there was a professor of physiology at Emory, who said that he believed he was about and he did to receive one of the first NIH contracts that was focused on this neural prosthesis program.

他提出要解答这个问题：如果我们从大脑多个位点进行记录，能否实时解析这些神经活动不仅用于控制肢体运动，还能仅凭神经活动预测肢体运动轨迹。

And he said he wanted to answer the question, if we were recording from a handful of places in the brain, would we be able to understand in real time how that neural activity was being used not only to move the limb, but could we then predict what the movement of the limb was from just the neural activity alone.

我至今记得坐在他办公室的情景——他名叫唐·汉弗莱——当他提出要利用神经活动来控制机械臂时，

And I remember sitting in his office, his name was Don Humphrey, as he said this, and I and he said he was gonna use the neural activity in order to control a robot arm.

我的第一反应是：什么？

And I remember what my response was, which was what?

你能再说一遍吗？

Can you say that again?

那一刻我清晰地意识到：只要我们对系统神经科学有足够认知，只要技术发展到能长期稳定记录大脑活动——当然是指单神经元集群层面的记录——这种可能性就真实存在。

And it was quite clear from that moment that there was a possibility that if we understood enough systems neuroscience, if the technology was developed well enough to be able to record chronically from the brain, we could record at the level of ensembles of single neurons, certainly.

如果我们能安全地实现这一点，并且这些记录能长期持续，还能实时解码这些活动，那么通过这类设备恢复丧失的神经功能这一希望确实可能成真。

And if we could do that safely, and those recordings lasted for a long period of time, and we could decode that activity in real time that this this hope that we'd be able to restore lost neurologic function with these devices could indeed come true.

这是个美好的起点，因为你知道，正是好奇心和我们对下一个前沿的兴奋感推动着人类不断前进。

That's a beautiful place to start because, you know, curiosity and us getting excited about the next horizon is what's keeping humanity moving forward.

而我们总是站在那个前沿的边缘，但那大概是什么时候的事？

And we're always at the edge of that horizon, but that was what year was that roughly?

1990年。

1990.

1990年。

1990.

这些对话就发生在那个时候。

That was when these conversations were happening.

我部分时间担任神经重症监护医生，在神经重症监护室工作。

So I spent some of my time as a neurocritical care doc, I work in the neuro ICU.

所以我见到的人们两天前还能自如活动、流畅说话，但昨天或今天早些时候突然遭遇了变故。

And so I see people who were able to move and were able to speak quite well two days ago, but yesterday, or earlier today, something happened.

脊髓损伤或脑干中风会导致患者陷入闭锁状态——清醒且警觉，却无法行动和言语，只能感知发生在自己身上及周围的一切，或许仅剩些许眼球运动能力作为唯一表达方式。

So spinal cord injury or a brainstem stroke that left that person locked in awake and alert, unable to move, unable to speak, but aware of everything that's happened to them and around them, except for perhaps some ability to move their eyes may be the only efforts.

我接触的肌萎缩侧索硬化症患者正逐渐丧失行动和语言能力。

I meet people with ALS who are slowly losing the ability to move and the ability to speak.

这个领域的研究重点（这很合理）一直致力于恢复这些患者更便捷的沟通能力和行动能力。

And a lot of the focus in this field, think rightly so, has been on trying to get to the point that we can restore any of those people's ability to communicate more easily and to move more easily.

我们的大脑步态研究已持续二十年。

Our brain gait studies have been going on now for twenty years.

这是一项关于植入式脑机接口的多中心研究。

It's a multisite study of a implanted brain computer interface.

在我们的研究及其他项目中，那些因ALS导致严重构音障碍的患者，通过植入不同类型的大脑记录传感器，我们得以解码与目标词汇相关的神经活动。

For our research and others, there are people who are profoundly dysarthric from ALS, who have different types of recording sensors placed in the brain, we're able to decode the neural activity associated with the intended words.

尽管由于严重构音障碍，患者实际发出的语音可能难以辨识，但神经活动可被解码——音素能被检测，词汇可被识别，最终通过计算机语音合成技术实现有声输出。

And even though those words when they're spoken may not be audibly intelligible because of the profound dysarthria, that neural activity can be decoded, the phonemes can be detected, the words can be recognized, and then through computer generated speech, those words can be heard aloud.

这是该领域近期取得的重要突破之一。

So that's one recent achievement for the field.

回顾过去二十年，已有案例显示无法移动手臂或手的患者仅通过想象自己手的动作就能控制电脑屏幕上的光标。

If we go back over the past twenty years, there have been examples of people unable to move their arms or hands who are controlling cursors on a computer screen just by thinking about the movement of their own hand.

也有针对颈髓损伤患者的演示，他们通过想象自己手的开合动作来实现控制。

There have been demonstrations of people also with cervical spinal cord injury, thinking about opening and closing their own hand.

他们之所以能做到这一点，不仅因为我们能检测到其运动意图，还因为能够通过刺激肌肉或周围神经来重新激活瘫痪的肢体。

And they're able to do so because not only can we detect the intended movement, but it's been possible to stimulate the muscles or peripheral nerves to reanimate that paralyzed limb.

该领域已有许多这样的示范案例，我认为这些确实推动了科学进步。

There's a number of these demonstrations in the field that I think have really been pushing the science forward.

就您个人研究而言，目前主要聚焦在哪些方面？

And in your case, your own particular research, where are you focusing on at this point?

是否有特别有趣的案例想与听众分享？

And are there any specific cases that are interesting that you would like to share with the audience?

是的，我们正在进行BrainGate临床试验——此刻我正坐在波士顿麻省总医院，这里是该试验的协调中心。

Yeah, so our brain gate clinical trials, I'm sitting at the moment at Mass General Hospital in Boston, which is the coordinating center for these trials.

我们与布朗大学、普罗维登斯退伍军人医疗中心、加州大学戴维斯分校、斯坦福大学和埃默里大学共同组成了BrainGate研究联盟。

Together with Brown University, the VA Health Center in Providence, University of California, Davis, Stanford and Emory, that's our BrainGate consortium.

我们正在进行这项临床试验，让居家患者使用实验性的BrainGate设备，他们要么尝试说话让我们解码其语音意图，要么尝试移动双手来控制平板电脑、机械臂或其他有助于日常生活的设备。

We're all engaged in this ongoing clinical trial where people who are at home have this investigational BrainGate device, and they're either attempting to speak while we decode their attempted speech or attempting to move their hands in order to gain control over either tablet computers or robot arms or other devices that might be helpful in their own daily lives.

我们最近发表在《新英格兰医学杂志》上的一篇论文，来自加州大学戴维斯分校的优秀团队，报道了一位四十多岁的ALS患者，他患有严重的构音障碍。

One of our most recent papers, which was in the New England Journal last year, from great brain gay teams out at the University of California Davis was reporting a gentleman in his forties with ALS who was truly profoundly dysarthric.

实际上只有他的妻子和护理伙伴能听懂他发出的模糊语音。

Only his wife and his care partner are really the only two people who hearing his audible speech can interpret it.

说实话，当我听到他说话时无法理解，研究团队也同样听不懂。

Certainly, I couldn't when I heard him and research team unable as well.

当然，其他任何想与他交流的人也都无法真正明白那些话语。

And certainly anybody else that might want to interact with him wouldn't really understand those words.

但他希望推动这项技术发展。

But he wanted to help to push this technology forward.

于是他参加了BrainGate试验，在其语言优势侧的中央前回运动皮层植入了电极阵列，这样不仅能记录单个神经元活动，还能捕捉到他试图说话时的神经元集群信号。

And he enrolled in the brain gate trial, he had arrays of electrodes placed into the motor cortex on his language dominant side that is placed into the dominant precentral gyrus, which allows for the recordings of not only single neurons, but ensembles of neurons associated with his attempt to speak.

在植入这些设备约三周后，我们进行了首次计划中的记录会话。

After placing these devices, about three weeks later was the first planned recording session.

他被要求读出屏幕上显示的词语。

He was presented and was asked to say some words that were presented on the screen.

这就是我们训练解码器或过滤器的方式，让它能够理解神经活动，并将这些神经活动转化回语言。

That's how we train the decoder or train the filter that understands the neural activity, if you will, and then translate that neural activity back into words.

经过约半小时的训练后，他被要求说出我们显示在屏幕上的短语。

And after about half an hour of training that computer program, he was asked to say some phrases that we had put on the screen.

他从约50个词的词汇库中选择词语尝试发音，这些词语随即在屏幕上显示。

Selecting from a vocabulary of about 50 words, he would attempt to say them and those words appeared on the screen.

这些词语通过非标准计算机合成语音呈现，而是使用了他ALS发病前的声音样本合成的深度仿造语音，这样听起来更像他本人。

And then they were made audible through a not necessarily standard computer generated speech, but through what's equivalently a deepfake for good, a version of his own voice pre onset of ALS that had been programmed so it would sound more like him.

你甚至能听到他声音中蕴含的情感。

And you could hear some of the emotion in his own voice.

据在场研究团队转述，当第一天系统成功运行时，所有人都因情绪激动不得不暂停工作。

And as the research team who was there relayed, everybody had to stop because of everybody's emotion when it worked on day one.

太不可思议了。

Amazing.

太神奇了。

Amazing.

而且，这已经足够震撼地展示了这类技术的能力。

And and that was striking enough in terms of a demonstration of what these types of technologies could do.

但仅有50个词汇量时，能表达的内容非常有限。

But with only a 50 word vocabulary, there's only so much that can be communicated.

他第二天休息了一天。

He got the next day off.

团队可没休息。

The team did not.

团队在周六加班加点，到了使用该设备的第二天大家又回来了。

The team worked hard over the Saturday, and then people came back on what was day two of using the device.

他获得或提供的词汇量扩展到了12.5万个单词，基本上涵盖了英语词典中除部分专有名词外的所有词汇。

He was presented with or he provided was with a 125,000 word vocabulary that he could select from, which is basically nearly all of the words except for perhaps some proper nouns in the English dictionary.

然后研究人员就让他随心所欲地说任何想说的话。

Then he was just asked to say whatever he wanted to say.

他说的第一句话是：这东西开着吗？

And the first thing he said was, is this thing on?

这句话不仅反映了技术确实在运作，还展现了他绝妙的幽默感。

Which kind of reflected not only the fact that indeed the technology was working, but his great sense of humor.

哦，天哪。

Oh my goodness.

太神奇了。

Amazing.

我有个问题。

I have a question.

这个设备，当你说他被要求说的时候，是指口头说出来，还是反映他应该想的内容？

This device, when you're saying he was asked to say, was it verbal say or was he was that reflective of what he was supposed to think?

我们知道机器在读取他的想法，但他是否也被要求实际说出那些话？

We know that the machine is reading what he's thinking, but he was he also asked to physically say that those words?

所以我不认为机器是在读取他的想法。

So I wouldn't say that the machine was reading what he was thinking.

实际上，我不会这么描述它。

Actually, that's not how I would describe it.

我们是从中央前回进行记录的。

We're recording from the precentral gyrus.

通俗地说，我们记录的是运动皮层，目标是仅解码尝试性言语。

Very generically, we're recording from motor cortex, and the goal is to only be decoding attempted speech.

所以我们无法知道他当时在想什么。

So we couldn't know what he was thinking about.

这是当前技术尚无法感知或解码的内容。

It's not something that the technology is capable of perceiving today or decoding.

但我们的目标是解码他实际尝试说出的言语。

But our goal is to decode his actual attempted speech.

就他的情况而言，至少在早期阶段，这是他真实的说话尝试，通过他构音障碍的声音可以听到。

And in his case, at least early on, it was his genuine attempt to speak, which is heard audibly through his dysarthric voice.

尽管如此，不仅在他尝试说话时有效，当他不尝试发声但仍有意向输出时也同样有效。

With that, though, not only does it work when he's attempting to speak, it also works when he's not attempting to vocalize, but he is still intending that output.

所以要准确描述这一点确实有点棘手，但即便是身体健全的人，也可以想象试图说些什么而不动嘴、不动舌头、不动声带。

So it gets a little bit tricky to describe this as well as we can, but even somebody who's able-bodied can think about attempting to say something without moving their mouth, without moving their tongue, without moving their vocal cords.

正是这种意图，这种动作可以被解码成屏幕上的文字。

And it's that intention, that action can be decoded into words on a screen.

太神奇了。

Amazing.

从尝试到解释或翻译是否存在延迟时间？

Is there a lag time between that attempt and the interpretation or translation thereof?

由于他在参与研究前约四年被诊断出患有ALS（肌萎缩侧索硬化症），并且正如言语病理学家常做的那样，他们曾指导他放慢语速以便他人能理解他的话。

So because he had been diagnosed with ALS about four years before joining the study, and as speech and language pathologists often do, they had instructed him to slow his speech down so that others might be able to understand what he's saying.

因此，由于这种训练和ALS本身的影响，他尝试性言语的速度约为每分钟30个单词——这大概是他能完整说出一句话的最快速度——我们以这个语速成功解码了他的尝试性言语，文字会逐个出现在屏幕上。

So as a result of that, and as a result of ALS itself, his attempted speech at pace was about 30 words per minute, that was probably about as fast as he could have attempted to just speak a sentence, for example, it was roughly at that 30 words per minute that we were able to decode his attempted speech, the words would appear on the screen one word at a time.

需要强调的是，这些文字之所以能出现，是因为首先检测到音素，然后通过统计预测这些音素更可能代表某个特定单词。

It is important to note that the words themselves appear because first phonemes are detected, and then those phonemes are assumed to be or predicted to be more statistically accurately representative of a particular word.

就像健全人打字或用笔记本电脑时看到的文字修正那样，当一连串单词组合后，系统会通过AI方式确保句子完全正确，而不仅仅是接近正确。

And then after a stream of words are put together, much like somebody who's able-bodied, whether they're texting or using their laptop might see word correction appear or might see an AI approach to getting that sentence actually right, as opposed to almost right.

这增加了一点额外的延迟。

That adds a little additional lag.

因此即使屏幕上每次出现最可能的单词，在结合上下文考虑后，这些单词中的一个很常见会被替换为第二可能的单词。

So that even if the most likely word one at a time appeared on screen, it's quite common that one of those words might have been replaced to the second most likely word after the rest of the sentences considered in context.

我们一直使用至今已有二十年的设备，这个设备是三十年前为基础科学研究开发的，我试着把它举到我的太阳穴附近，我会把它靠近摄像头，我们看看效果如何。

The device that we've been using and have now been using for twenty years, and it was a device that was developed for basic science research thirty years ago, I'm going try to hold this up against my picky, and I'll bring it close to the camera, and we'll see how well this works.

这是一个4×4毫米的平台。

That is a four by four millimeter platform.

如果你眯起眼睛看，而且我刚好把它对焦清楚，这个4×4毫米的平台上分布着100个电极。

If you squint your eyes, and I happen to get it in focus, there are 100 electrodes on that four by four millimeter platform.

每个电极长度1.5毫米，可以放置在大脑表面并接入大脑，让我们能够探测到大脑皮层的浅表层，大概到三五边界区域。

They're each 1.5 millimeters in length, and it can be placed on top of the brain and tapped into the brain, allowing us to get to the superficial layers of cortex, maybe to the three five border.

每个电极都有可能记录到一两个单独的神经元活动，虽然并非所有电极都能记录到神经元信号，有些会记录到一两个。

And each of those electrodes has the potential to be able to record from one maybe two individual neurons, not all of them will necessarily record from a neuron, some will record one or two.

它还能记录多个神经元在电极尖端附近的综合活动，并能记录一些局部场电位——这些是能从大脑内部记录到的更类似模拟信号的活动。

It also can record the sum activity from multiple neurons near any of those electrode tips, and can record some of the local field potentials, the more analog activity that can be recorded from inside the brain.

我刚才展示的那个设备叫做犹他电极阵列。

That device that I showed you is called the Utah Array.

它是由犹他大学的迪克·诺曼在三十多年前开发的。

It was developed by Dick Norman at the University of Utah back thirty plus years ago now.

至今仍被用于我们的脑门研究项目。

It continues to be used in our brain gate research.

太神奇了。

Amazing.

你认为这个领域未来几年会如何发展？特别是随着人工智能、机器学习以及所有这些增强元素的加入？

Where do you see this field moving in the next few years, especially with the addition of AI and machine learning and all these augmented elements?

你觉得未来两三年内这个领域会走向何方？

Where do you see it going in the next couple of years?

是的，这个领域最初是从解码部分开始的，计算神经科学这部分真正起源于我们对单个神经元与意向运动关系的基础理解——具体来说，就是布罗德曼第五层初级运动皮层的皮质运动神经元，我们知道这个细胞很可能与脊髓运动神经元相连。

Yeah, so the field started with the decoding part of the field, the computational neuroscience part of this field really started with our fundamental understanding of how individual neurons are related to intended movement, that is an individual neuron, a corticomotor neuronal cell in layer five of Brodmann for primary motor cortex, we know that cell is likely connected to a motor neuron spinal cord.

因此当那个神经元放电时，它实际上与肌肉活动——或者多块肌肉的活动相关。

And so when that neuron fires, it is really related to a muscle activity, or maybe the activity of a muscles.

过去五年里，机器学习和人工智能的快速引入、普及及其强大能力，使我们不再需要假设神经活动的具体作用，而是可以采取假设无关的方式，直接将神经活动与其目标相匹配，并创建解码方法——这确实推动了整个领域向前发展。

The past five years, the rapid entrance and availability and the power of machine learning of AI to not necessarily have a hypothesis for what that neural activity is doing, but to be hypothesis agnostic, and to match the activity with its goal, and then to create the decoding approach that's really moved the field forward.

这也是我们在过去几年中看到某些技术进步的部分原因。

And it's part of the reason for some of the advances we've seen in just the past few years.

我认为只需短短几年时间，市场上就会出现用于交流或行动功能的植入式脑机接口设备。

I think it's just a handful of years before we actually have devices available on the market that are implanted BCIs for communication or mobility.

需要说明的是，目前已有用于癫痫治疗的响应性神经刺激技术，这完全可以被视为一种植入式脑机接口。

I should say that there are already something that would quite reasonably be considered an implanted BCI in responsive neural stimulation for epilepsy.

这些设备记录神经活动，目的是检测癫痫发作，如果检测到临床前期的癫痫发作，就会进行刺激以抑制发作，防止其发展为临床发作。这至少初步构成了一个脑机接口。

Those devices record neural activity, the goal is to detect seizure, and if detected, if detecting a preclinical seizure, then to stimulate and to suppress that seizure before it becomes clinically That, at least at a first pass, is a BCI.

它记录神经活动，解读信息，并利用这些信息做出有益的行动。

It records neural activity, it interprets the information, and it does something useful with it.

这真是太棒了。

That's wonderful.

在这个领域中，你们需要应对哪些伦理问题？

What are the ethical issues that you deal with in this realm?

这涉及到自主权问题、隐私问题以及长期护理问题。

There's questions of autonomy, there's questions of privacy, long term care.

当有人说他们是在没有受到任何胁迫的情况下自主做出决定时，考虑到他们所处的环境，实际情况并不像人们想象的那么明确。

And when somebody says that they're making a decision on their own and without any coercion, that's not as clear cut as people think it is, given the situations they're in.

我很想了解您认为在这个领域及其发展过程中，我们需要应对哪些伦理问题。

I would love to know what you think are the ethical issues that we need to deal with this field and where it's going.

是的，确实存在几个非常重要的伦理问题。

Yeah, there are several and they're really important.

在我们自己的脑机接口研究中，自二十年前启动试验以来，我们就一直在深入思考这些问题。

And in our own brain gate research, we've been thinking about these deeply since we launched the trial twenty years ago.

其中一个问题是我们在进行大脑记录。

One of them is we're recording from the brain.

这就是植入式脑机接口的起点。

That's that's where the implantable BCI starts.

而且我们记录的信息也涉及隐私。

And there's information that we're recording.

我主张我们记录的信息归属于被记录者，你的大脑数据属于你自己。

And I would propose that information that we're recording belongs to the person from whom we're recording, that your brain data are yours.

这看似简单——为了使其有用，我们进行的记录需要被解码并以某种实用方式反馈，无论是控制电脑屏幕上的光标、操控肢体，还是重建语音。

It seems like a simple thing to say those data in order to be useful, the recording that's being made is being made so that it can then be decoded and returned in some useful way, whether it's for control of a cursor on a computer screen or control over to somebody's own limb, or recreated speech.

我和许多人都非常认同：这些神经数据只能在数据生成者许可的情况下与他人共享。

I and I think many are very comfortable with the fact that those neural data should only be shared with anybody else with permission of the person who generated the neural data in the first place.

如果我们沿着这个思路，设想一个完全通过审批的市场化设备正在被使用，这些神经数据的所有权归谁？

If we head down the line and we think about a fully cleared marketed device that somebody is using, who is the owner of those neural data?

是个人吗？

Is it the person?

是临床医生吗？

Is it the clinician?

是公司吗？

Is it the company?

我们记录的数据内容究竟是什么？

What is the content of the data that we're recording?

当我在记录时，当我们的团队在记录一个意图手部动作时？

When I'm recording, when our team is recording an intended hand movement?

这正是我们在此的目的——试图让某人能够解码或控制电脑屏幕上的光标，举例来说。

That's the reason that we're there is to try to allow somebody to decode or somebody to control a cursor on a computer screen, for example.

但毫无疑问，这些神经活动中还蕴含着其他信息，比如他们当时的警觉程度，这些似乎也应包含在我们记录的神经活动中。而要就这些设备可能记录的内容提供充分知情同意，确实颇具挑战。

But there's undoubtedly other information in that neural activity could be reflecting, for example, their level of alertness at that moment seems like that should be somewhere in the neural activity that we're recording, providing the best informed consent that we can about what these devices may be recording is challenging.

我们尽力分享这些信息，也坦诚认知的局限，但我认为这引发了一些伦理问题。

We share this information as best we can, we share the limits of our knowledge, but I think that raises some ethical questions.

还有一点，获取这些技术的途径至关重要。

There are others, the access to these technologies is incredibly important.

这与任何其他先进医疗技术并无二致。

This is not different than any other advanced medical technology.

但我真心希望这些技术能广泛惠及所有可能从中受益的人群。

But these technologies, I really hope are going to be widely accessible to anybody who would benefit from them.

作为由医生、科学家、瘫痪亲历者、监管者和费用支付方组成的共同体，我们必须共同解决如何让这些技术真正普及，并确保其持续发展以推出第二代产品。

As a community of physicians and scientists and people with lived experience of paralysis and regulators and people responsible for reimbursement, I think as a community, we really have to figure out how will these technologies actually get out there and get out there widely and be available in a manner that allows the technology to continue to develop so that version two can be created.

我认为不仅需要开展这些对话，更重要的是让这些对话能够影响设备研发过程中的一些决策。

I think it's important not only have these conversations, but to allow these conversations inform some of the decisions that are made during the device development.

对于植入式脑机接口，我们理论上可以使用各种植入式硬件加密技术来确保数据传输的绝对安全，但这需要消耗电力。

For an implanted brain computer interface, we could potentially use all kinds of implanted hardware encryption to be absolutely sure that the data that are transmitted securely, that takes power.

这些芯片需要配备电池，它们需要某种形式的电源支持。

Those chips require a battery, they require power of some sort.

目前可用的电力预算相当有限。

There's a pretty limited power budget.

高性能芯片也会发热。

Chips that are powerful also become hot.

而我们不希望植入设备产生热量。

And we don't like heat in implanted devices.

事实上，对于植入或靠近神经系统的设备，其周围产生的热量有非常严格的限制。

In fact, there's a very strict limit to how much heat can be generated around a device that's placed in our near the nervous system.

但如果传输这些数据的目的只是在一米范围内传送，且超过一米就无法接收，而且传输的数据只是用户控制电脑发送推文的指令，那么从一开始就没有加密的必要。

But if the goal of transmitting those data was only to transmit it a meter away anyway, and nothing beyond a meter could pick it up, and the data that are being transmitted are again, somebody's requests for controlling the computer so they can send out a tweet, there was no reason to encrypt it in the first place.

因此需要在技术能为人们带来的益处与强加的安全需求之间找到平衡

So there's a balance between the benefit that technology could provide somebody, which may be somewhat tempered by the security requirements that are imposed

在

on

它上面。

it.

完全同意。

Absolutely.

你认为五十年后我们会发展到什么程度？

Where do you see us in fifty years?

说实话，在我看来，以目前的发展速度，五十年都太长了。

Or to be honest, fifty years is too long at the rate things are moving from my perspective.

那么在二十到二十五年内，你认为瘫痪患者中会有相当大比例的人能够重新行走吗？特别是借助人工智能、增强数据、实时数据处理和实施技术？

And twenty to twenty five years, do you see that people who have paralysis, a great number of them would be able to walk, especially with AI and augmented data, real time data processing and implementation?

既然你好心地给了我五十年的时间跨度，这很棒。

Since you were kind enough to give me a fifty year runway, which is great.

我认为你刚才提到的许多目标在那个时间段内完全有可能实现，仅看过去五年发生的变化就可见一斑。

I think a lot of the things that you just said are well within reach in that period of time, just seeing what's happened over the past five years.

你描述的下肢运动功能恢复方面，我们已经看到一些成功案例——通过相对简单的脊髓刺激技术帮助患者恢复基础下肢运动。若结合皮层或其他脑部信号记录来驱动脊髓刺激，还能进一步改善功能，实现更复杂的下肢活动，比如调节步态节奏。不过躯干稳定性等需求可能超出了脑机接口的必要范畴。

Some of what you described lower extremity mobility, we've seen some nice examples of some pretty straightforward spinal cord stimulation being used to re enable some lower extremity movement that may be further improved by using cortical or other cerebral recording in order to drive some of that spinal stimulation to allow for even more complex lower extremity activity to change the pace of walking or there's a lot of trunk stability that is required that's outside necessarily the realm of a brain computer interface.

随着神经调控和神经记录技术各自的发展，我认为皮层内微刺激技术的进步将带来某种形式的视力改善。

As these neuromodulation and neural recording technologies independently advance, I think we'll see some improvements in intracortical microstimulation leading to some version of improved vision.

在更短的时间内，我希望我们能向渐冻症患者承诺：他们永远不会丧失沟通能力。

With a much shorter runway, what I hope is that we're gonna be able to tell somebody with ALS that they will never lose the ability to communicate.

毫无疑问，我们需要神经学界的全力参与。

What's required is without a doubt, we need the engagement of the neurology community.

今天能与您合作令我倍感振奋，但我们需要更多神经科医生加入。

I'm thrilled to be joining you today, but we need more neurologists.

希望在您的倡导和平台支持下，我们能达成这个目标。

And hopefully, with your advocacy and platform, we can get there.

我认为神经科医生最有能力判断哪些姑息性设备最适合诊所里的具体患者，或是推荐两三种方案，而不必考虑第四第五种选择。

I think neurologists are well positioned to understand of the hopefully palliative devices that will be available, which ones are most likely to benefit any individual patient in their clinic, or to make recommendations to consider two or three of them, but not necessarily the fourth or fifth.

我认为这将是必需的，并且显然需要持续获得慈善和联邦资金支持，以便研究能够继续创造出更好的设备。

I think that will be required and clearly continued philanthropic and federal funding so that the research can continue to create ever better devices.

我们非常感谢博士。

We're so thankful to Doctor.

霍克伯格的深刻见解，特别是在医疗平等方面的观点。

Hockberg for his insights, especially when it comes to the equality of care.

尽管这些优质资源通常会被引入社区，但平等性和可及性却完全没有得到解决。

And even though these fantastic resources are usually introduced into the community, but equality and access is really not addressed at all.

往往最需要帮助的人群反而最难获得这些资源。

It often doesn't get to the people that need it the most.

是的。

Yeah.

这非常令人遗憾，我认为这已经成为一种历史模式——最脆弱的群体和最需要这些工具的人往往最后才能获得帮助。

And that's very sad, and I think that has been the historical pattern where the most vulnerable people and the ones that need the tools the most are often the last to get it.

这几乎造成了一种资源分配的不平衡——那些有经济能力或身处临床试验研究中心附近的人更容易获得帮助，但事情本不该如此。

It almost becomes an imbalance of providing this for people who may have the means or who are present around that clinical trial in the research center, but it doesn't have to be that way.

正如你们稍后将听到下一位嘉宾布兰德蒙德博士和斯塔维斯基博士所言，目前这个领域仍处于前竞争和商业化前期阶段，这实际上让研究人员、监管机构和公众都能对此发表意见。

And as you will hear in a moment from our next guests, doctor Brandmond and doctor Staviski, this is still a precompetitive and a precommercial moment in the field, and it actually allows the researchers, the regulators, the public to actually have a say in this.

他们可以围坐在同一张桌子旁，就安全、安全、可及性等重要议题展开对话

And they can sit around the same table and have conversations about important things such as safety, safety, access

伦理问题。

The ethics.

所有这些的伦理考量。

The ethics of all of it.

甚至那些对其技术细节并不真正了解的人，也应该有权决定谁可以获得它、如何获得它，以及如何让所有人都能使用它。

And even people who are not really aware of the specific tech part of it should have a say in determining who gets it, how do they get it, and how it should be available for everyone.

对此我持谨慎乐观态度。

I'm cautiously optimistic about that.

这一切将我们引向与大卫·布兰德曼博士和谢尔盖·斯塔维斯基博士的对话，他们作为加州大学戴维斯分校神经假体实验室的联合主任，正工作在工程、创新和以患者为中心护理的交叉领域。

And all of this brings us to our conversation with doctor David Brandman and doctor Sergey Staviski who are working right at the intersection of engineering, innovation, and patient centered care as co directors of the neuroprosthetic lab at UC Davis.

谢尔盖和大卫正在构建系统、合作伙伴关系和开放框架，使符合伦理、可扩展且包容的脑机接口部署成为可能。

Sergei and David are building the systems, the partnerships, and open frameworks that can make ethical, scalable, and inclusive deployments of brain computer interface possible.

大卫，你是怎么踏上这条探索之路的？

David, how did you get into this journey?

是的。

Yeah.

谢谢提问。

Thanks for the question.

我是一名神经外科医生。

I'm a neurosurgeon.

天啊，我是怎么进入脑机接口领域的呢？

And gosh, how did I get into brain computer interfaces?

2010年我在加拿大新斯科舍省哈利法克斯市进入神经外科住院医师培训，开始实习不到一年就意识到自己渴望学术发展道路。

When I matched to neurosurgical residency in Halifax, Nova Scotia in 2010, within a year of starting my residency, I realized I wanted an academic trajectory.

我喜欢挑战临床医学认知边界并帮助他人的理念。

Like I liked the idea of pushing the boundaries of what we know in clinical medicine and helping folks out.

于是在2011年，我参加了一个工程学会议，遇到一位做脑机接口演讲的学者，当时就觉得这简直是世界上最酷的东西。

So in 2011, I went to an engineering conference and I met a guy who was giving a talk on brain computer interfaces and I said, this is the coolest thing ever.

我走上前对他说，你好，我是一名神经外科住院医生。

I walked up to him and I said, hi, I'm a neurosurgical resident.

你能多告诉我一些关于脑机接口的事吗？

Can you tell me more about BCI?

于是他请我喝了杯咖啡，简单介绍了这个领域，两年后我就成了他的博士生。

So he bought me a coffee and tell me a little about the field and then he fast forward two years later and I'm his PhD student.

2011年我在一场会议上遇到了李·霍克伯格，当时他正在介绍自己参与的BrindGate二期临床试验工作，后来我于2013至2017年加入他的课题组攻读博士学位。

So I met Lee Hochberg at a conference as he was talking about his work as part of the BrinGate two clinical trial back in 2011 and actually joined his group as a PhD student from 2013 to 2017.

我做了个短暂的'胜利巡游'——这是我对那段短暂博士后经历的戏称，之后继续完成住院医师培训，接着在埃默里大学跟随尼克·布利斯、鲍勃·格罗斯和尼克·杨开始功能性立体定向研究。

I did a short victory lap as I like to call it, doing a very short postdoc before continuing to finish residency and then starting to do my functional stereotactic fellowship at Emory University under Nick Bullis, Bob Gross and Nickel Young.

2021年我以教职身份加入加州大学戴维斯分校。

I started out in faculty in UC Davis in 2021.

所以对我来说，脑机接口是顺理成章的选择。

So for me, BCI made a lot of sense.

我的专业背景是物理学。

My background was in physics.

我一直很喜欢计算机编程，并且看到了需求。

I've always enjoyed computer programming and I saw a need.

就像我在神经外科培训期间遇到的瘫痪病人，他们渴望与环境互动却无能为力。

Like I met patients as part of my neurosurgical training that were paralyzed, that wanted to interact with their environment but couldn't.

我认为脑机接口能完美结合我的专长与兴趣，同时直接帮助有需要的人。

And I saw BCIs as a really attractive way to build on my strengths and my interests while still being able to directly help people in need.

太棒了。

Amazing.

简直不可思议。

Absolutely amazing.

我很欣赏你能找到这个方向。

I love that you were able to get that direction.

我想很多神经科医生、神经外科医生和神经科学家都有这种远见，知道工程学与生物学的革命将在两界交汇处发生。

I think a lot of the neurologists and neurosurgeons and neuroscientists had that in mind, knowing that there's going to be an interface between these two realms of the revolution of engineering and biology.

你们成功找到了这个交汇点。

You guys were able to find it.

谢尔盖，你那精彩的经历是怎样的？

Sergei, what was your amazing journey?

是啊。

Yeah.

你会听到一些相似的主题。

You're gonna hear some similar themes.

我在布朗大学读本科时，2004年入学，恰好是原始BrainGate系统——首个采用多电极阵列的慢性脑机接口——正由布朗大学衍生的一家初创公司进行测试的年份。

So I was a undergraduate at Brown University, and I started college in 2004, which happened to be the year that the original BrainGate system, which is the first chronic brain computer interface using a multielectrode array, was actually being tested by a spinout company that came out of Brown.

于是我就去上大学了。

And so I went to college.

我一直对工程学很感兴趣。

I'd always been interested in engineering.

我热爱编程。

I love to program.

我喜欢动手制作东西。

I like to build things.

我对生命科学有着广泛的兴趣。

I was interested in life sciences broadly.

从哲学角度，我对心智也很感兴趣，就像大多数18岁、有点自恋的年轻人一样，想了解自己以及是什么造就了现在的我。

I was philosophically interested in the mind as well as I think most 18 year olds who are a little bit narcissistic and wanna know about themselves and what makes me who I am.

当时我正在上神经科学导论课，不知为何在布朗大学，这实际上是你秋季学期上的第一门生物课。

And I was taking this intro to neuroscience class, which for whatever reason at Brown is actually the first biology class you take in the fall.

另一位教授约翰·多纳休（与李·霍克伯格密切合作）客座讲授了脑机接口的概念，并展示了他们与首位参与Cyberkinetics脑机接口试验的受试者刚出炉的最新成果视频。

And a different professor, John Donahue, who actually worked very closely with Lee Hockberg, gave one of the guest lectures talking about this idea of a brain computer interface and showing some videos from their, like, fresh, hot off the presses results with their very first participant in this cyberkinetics brain gate trial.

那展示的是一位比我年长几岁的年轻男子，他在一场刀战中受伤，导致颈部以下瘫痪。

And that was showing a young man, just a few years older than me, who had been injured in a knife fight and was paralyzed from the neck down.

在这首次真正惊人的成果中，几年后发表在《自然》杂志上，引起了巨大轰动，但本质上，他当时正试图移动自己的手。

And in this first ever really phenomenal result, it was published a couple years later in Nature, was a big deal, but basically, he was trying to move his hand.

他们正在解码他试图移动手时的神经活动，他当时正利用这些信号——以今天的标准来看——非常不精确地控制电脑光标。

They were decoding the neural activity as he was trying to move his hand, and he was using that to, by today's standards, very poorly control a computer cursor.

但尽管如此，他确实能够移动那个电脑光标，甚至还能操控一个简单的机械手臂。

But nonetheless, he was able to move that computer cursor and even had to move a little simple robotic arm.

我觉得那简直酷毙了。

I thought that was the coolest thing ever.

我当时就说：这太神奇了，而且完全符合我的兴趣。

I said, this is amazing, and this perfectly aligns.

我一直想做些与医疗相关的事情来帮助他人。

I wanted to do something medical related to help people.

于是我加入了BrainGate项目——那时它已转型为学术研究而非商业项目，之后我又去斯坦福攻读博士学位。

So I worked for BrainGate, which at that point had become an academic endeavor rather than a commercial one, and then went on to do my PhD at Stanford.

实际上我和大卫有过短暂交集，后来我们同一天入职加州大学戴维斯分校担任助理教授。

Actually, I overlapped briefly with David and then started my assistant professor job here at UC Davis on the same day as David.

这倒不是刻意安排的同步招聘，但也不是纯属巧合。

It was not like a super coordinated hire, but it was not coincidence.

简单来说，我们俩同时申请了加州大学戴维斯分校同一系里的两个不同职位，当时的系主任发现：『我这里有个想做脑机接口的神经外科医生，还有个想做脑机接口的神经科学家』。

So basically, both of us applied for two separate jobs in the same department at UC Davis, and our chair at the time said, I've got a neurosurgeon and a neuroscientist that both want to do brain computer interfaces.

他们真该好好聊聊。

They should probably talk.

我们确实这么做了，并且很快达成共识，发现彼此兴趣非常相似，决定共同运营一个实验室。

And we did, and we quickly agreed that we had very similar interests and wanted to run a lab together.

所以就有了现在的我们。

So here we are.

太棒了。

Amazing.

我认为我们正站在重大突破的边缘。

Right now I think we're at the precipice of something huge.

能告诉我们你们目前的研究进展和未来方向吗？

So tell us where you are in that journey and where are going?

作为开场白，我先介绍一下BrainGate的临床情况。这是一项临床试验，志愿者通过脑部手术在大脑控制运动的区域植入电极。

So maybe by way of introduction, I'll just tell you about the clinical So BrainGate, it's a clinical trial in which people volunteer to undergo brain surgery to place electrodes into part of the brain that controls movement.

这项临床试验的主要目的是研究我们植入的电极是否安全。

So the point of the clinical trial is to study whether or not the electrodes we're implanting are safe.

但实际上，参与临床试验的志愿者还会与科研团队合作，共同开发新一代神经技术来帮助瘫痪患者。

But actually as a consequence of signing up for the clinical trial they work with a scientific team to develop next generation neurotechnologies to help people with paralysis.

这项临床试验目前正在进行中，由麻省总医院主导，属于研究者发起的临床试验，即由个人而非公司负责运营。

So the clinical trial is ongoing, it's led out of Mass General Hospital and it's a sponsor investigator led clinical trial which means a guy runs the clinical trial as opposed to a company running the clinical trial.

那个人就是李·霍克伯格。

That guy is Lee Hochberg.

基本上，该试验面向因脊髓损伤、脑干中风、肌萎缩侧索硬化症及其他神经退行性疾病导致瘫痪的患者开放。

Essentially it's open to people living with paralysis from things like spinal cord injury, brainstem stroke, ALS and other kinds of neurodegenerative diseases.

作为临床试验的一部分，他们会接受这些电极的手术植入，随后在一年时间内与科学家和工程师合作开发神经假体。

And as part of the clinical trial they undergo surgical implantation of these electrodes and then over the course of a year, with scientists and engineers to develop neuroprosthetics.

可以说，诊所成立的前二十年，确切地说是临床试验的前十七年，我们主要致力于通过点选鼠标控制来恢复交流能力，或是通过机械臂控制来恢复功能独立性。

So I would say that the first twenty years of the clinic, the first seventeen years of the clinical trial really focused on restoring communication by way of point and click mouse control or restoring functional independence by way of robotic arm control.

具体来说，就像你操作电脑时用手移动鼠标，光标随之在屏幕上移动并点击目标那样。

So what I mean by that is as you interact with your computer, you take your hand and you move your mouse and then your mouse moves on a computer and you can click on things.

但如果你瘫痪了，就无法完成这些动作。

But if you're paralyzed, you can't do that.

因此我们在前十七年至今仍在探索的是：能否仅凭思维来操控电脑光标？

And so what we were learning for the first seventeen years and still learning right now, can you control a computer cursor using your thoughts?