大脑如何识别错误：与Camille Gontier博士探讨BCI中的内源性错误检测

我们在斯特拉斯堡所做的研究截然不同，我们将进行闭环神经调控以应用于精神疾病治疗。

So what we are doing here in Strasbourg is, it's widely different, we are going to do closed loop neuron modulation for application to psychiatric disorders.

这其中有几个不同之处。

So there are several differences.

首先是初期主要聚焦于非侵入性技术。

The first one is that it's at the beginning at first, mostly going to focus on non invasive technologies.

即记录和刺激都将采用非侵入性技术。

So non invasive technologies for both recordings and stimulation.

也就是说，记录将通过脑电图（EEG）完成。

That is to say, of the recordings is going to be done with EEG.

而大多数其他刺激将通过经颅磁刺激（TMS）或电流刺激实现。

Most of the other stimulations are going to be done with TMS or current stimulations.

因此现阶段仅涉及非侵入性手段。

So for the moment, only non invasive stuff.

嗯。

Mhmm.

第二个不同之处在于，与匹兹堡团队专注于运动解码或感觉刺激不同，我们将重点研究针对精神疾病的神经调控技术。

The second difference is that instead of focusing on motor decoding or sensory stimulations as people in Pittsburgh are doing, we are going to focus on neuromodulation for psychiatric disorders.

已有实证研究表明，对大脑进行电或磁刺激能够改善症状并缓解某些病理状态。

It has been empirically shown that electrical or magnetic stimulation of the brain can yield improvement in symptoms and in some pathologies.

我认为最广为人知的例子就是帕金森病的深部脑刺激疗法。

I think I would say the most widely known example is deep brain stimulation for Parkinson's disease.

将电极植入患者大脑后施加刺激，虽然不能完全消除但能显著减少震颤并改善步态运动，这效果非常令人印象深刻。

You put an electrode into the brain of the patient, you elicit some stimulation and probably not no more, but reduced tremors and improved gait movement, which is super impressive.

还有一系列研究正在尝试探索不同类型的调控是否也能对其他类型的障碍产生影响。

And there is a line of research trying to see whether different kind of modulation can also have effects on other kinds of disorders.

比如我在想，能否通过刺激大脑活动来增强或抑制脑电图某些频段的活性，从而缓解某些症状，例如阿尔茨海默症或多动症。

I am thinking, for instance, about being able to stimulate brain activity to increase or reduce the activity into some frequency bands of the EEG, to be able to solve some symptoms, for instance, for Alzheimer's disease or for ADHD.

因此整个研究方向都致力于神经调控技术及其在精神疾病中的应用。

So there is a whole line of research trying to do neuromodulation and to apply it to psychiatric disorders.

我认为目前的主要问题在于这种方法还非常依赖经验性探索。

I would say that the main issue at the moment is that it's very empirical.

目前缺乏明确的量化依据，无法确定何种刺激会引发何种脑部活动，以及通过何种方式改善哪些症状。

There is no clear quantitative insights as to what kind of stimulation is going to elicit what kind of activity and improve which symptoms in which way.

例如，在深部脑刺激领域，目前很大程度上仍依赖经验性操作。

For instance, in deep brain stimulation, it is still pretty much empirical.

将电极置于特定位置，施加某种刺激模式，就能获得某种改善效果。

You put an electrode at some location and you apply some kind of pattern of stimulation and you obtain some kind of improvement.

因此我们的目标是实现高效的闭环神经调控，即能够实时优化刺激方案。

So what we would try to do is to have efficient closed loop neuromodulation in the sense that we would like to have on the fly improvement of the stimulation.

也就是说在每个时间步骤中，能够检测前序刺激产生的效果，并实时调整以达到预期目标。

That is to say at each time step, being able to detect the effects that stimulation at previous time steps had, and to be able to refine them on the fly to converge to the desired effect.

这是一个截然不同的研究领域，仍处于开放探索的初级阶段。

So a widely different field of research, still very open, still in its infancy.

我认为这仍属于PCI范畴，因为它仍涉及对某些心理状态的解码和闭环式的针对性刺激。

I would say it's still considered to be PCI because it still involves decoding of some mental states and ad hoc stimulations in a closed loop manner.

最令人兴奋的是，这个领域将融合极具价值的临床应用和激动人心的理论挑战。

And what is going to be exciting about that is that it's going to mix both very interesting clinical applications and also very exciting theoretical challenges.

我们需要能够在计算机模拟中刺激大脑的效果，并且需要能够进行主动学习以闭合循环，从而实时优化。

We need to be able to stimulate in silico the effect of stimulations on the brain, and we need to be able to perform active learning to close the loop and to be able to optimize it on the fly.

所以我对此非常期待。

So I'm really looking forward to it.

是的。

Yeah.

没错。

Yeah.

这听起来确实令人兴奋。

It sounds absolutely exciting.

而且我们知道最近神经刺激及其应用正在兴起。

And yes, we know that lately we have a rise of neurostimulation and its applications.

正如你所说，针对阿尔茨海默病、一般痴呆症患者改善其认知功能，当然还有神经发育障碍患者。

And like you said, for Alzheimer's disease, patients with dementia in general for improvement of their cognitive functions and of course people with neurodevelopmental disorders.

因此有这么多惊人的应用，我认为这个领域才刚刚起步。

And so there are so many amazing applications and the field I think is just starting.

是的，这里正是合适的地方。

Yes, it's the right place to be here.

我的问题是，你从匹兹堡大学当前项目或已完成项目中学到的哪些技能和知识可以应用于新工作？

My question is what skills, what things learned from your current project at the University of Pittsburgh or the project that you finished, you can use for your new work?

在你已掌握的内容中，有哪些可以应用于精神障碍的神经调控研究？

Anything from what you already learned during the process you can apply in your work on this neuromodulation for psychiatric disorders?

我想我的回答会非常简短且无趣。

I think my answer is going to be very brief and not interesting.

我并不完全确定。

I'm not absolutely sure.

可以肯定的是，在匹兹堡Jen Collinger实验室做博士后期间，我学到了大量关于脑机接口的知识，了解不同方法的工作原理。

So for sure, during this postdoc in Pittsburgh in the lab of Jen Collinger, I learned an insane amount of stuff about BCI, about about how different methods work.

当我说方法时，既包括硬件也包括软件。

And when I say methods, mean both hardware and software.

进行分类的最佳方式是什么？

What are the best way to do classification?

进行解码的最佳方法是什么？

What are the best way to do decoding?

就在我们交谈的同时，匹兹堡大学与瑞士伯尔尼大学之间还有一个正在进行中的合作项目——我在伯尔尼大学完成了博士学位——匹兹堡大学负责部分数据采集工作，而伯尔尼大学的理论研究团队正在尝试各种新的非线性方法进行运动解码，以验证我们是否能改进现有技术。

As we speak, we still have an ongoing project between the University of Pittsburgh and the University of Bern in Switzerland, in which I did my PhDs, in which the University of Pittsburgh is doing some recordings, and the theoretical people at the University of Bern are playing around with different new nonlinear methods to do motor decoding to see whether we can improve previous methods.

但目前而言，我在匹兹堡积累的庞大知识体系中哪些具体内容会对这些神经调控项目特别有用，我仍不确定。

But at the moment, which specific elements of this huge background of stuff that I've been able to learn in Pittsburgh are going to be especially useful for these neuromodulation projects, I'm still not sure.

纯粹是因为这个项目几周前才刚启动，最终研究方向尚未明确。

Simply because this is a this is a project that I just started just just a few weeks ago, and it's still not clear in which direction we are going to go.

但可以确定的是，我清楚地知道哪些方法可用于心理状态解码——无论是线性方法还是非线性方法各有利弊，更重要的是掌握了与人类受试者合作的研究经验。

But what is clear is that we I know I have a clear idea of what kind of methods can be used for decoding of mental states, whether it's better to use linear methods or nonlinear methods, each method has its perks and its inconvenience, and especially what it's like to work with with human participants.

可以说在匹兹堡的收获确实非常丰富。

So I definitely learned a lot in Pittsburgh.

只是不确定所有这些知识要素将如何融合并作用于这个新项目。

I'm not sure what and all these elements are going to converge and contribute to this new project.

核心观点是：无论是跟随Pfeister教授攻读博士期间，还是与Jen Collinger进行博士后研究时积累的经验，都将成为无价之宝。

The main idea is that everything that I've been learning, either with professor Pfeister during my PhD or with Jen Collinger during my postdocs, are going to be invaluable.

我想说的是，有很多软技能，比如能够高效地思考不同方法，虽然难以言表，但这些绝对会对我未来的研究生涯大有裨益。

And I would say there are a lot of soft skills like being able to think efficiently about different methods, which it's hard to put words on, but it's definitely going to be useful for the rest of my career in research.

是的。

Yeah.

这不仅关乎技能，更关乎思维方式。

It's also not only about skills, it's about the mindset.

对。

Yes.

你如何看待事物。

How you think about things.

没错。

Yes.

非常感谢你提到这一点。

And thank you very much for noting that.

嗯。

Yeah.

这非常重要。

It is very important.

既然我们是一档职业导向的播客，我想就你的职业道路请教几个问题。

And as we have a career oriented podcast, I would like to ask you several questions about your career path.

我们看到当前趋势是人们选择进入工业界而非留在学术界。

We see the trend of people going to the industry instead of staying in academia.

而你实际上选择留在学术界。

And you are actually staying in academia.

是的，你完成了博士后研究，现在又获得了新职位，将继续开发这些方法——没错，是针对神经调控和精神障碍的方法。

Yes, you finished your postdoc and now you have another position in which you will be working on developing these methods, yes, for neuromodulation, for psychiatric disorders.

是什么促使你决定继续走这条学术道路呢？

What makes you want to continue on this academic path?

哦，我觉得这是个很好的问题。

Oh, I say I would say it's a good question.

或许我应该先说明，法国的学术体系与其他国家存在很大差异。

Maybe I should start by saying that the French system of academia is widely different from what exists in different countries.

我想说在大多数国家，包括美国或瑞士等，获得终身教职并留在学术界非常困难，因为博士后数量远超过教授职位，想要留在学术界确实很不容易。

I would say that in most countries, including The US or Switzerland, for instance, it's very hard to find a tenured position and to remain in academia once you have done your your postdocs, simply because you have way more postdocs than tenured professor positions, and it can be really hard to stay in academia.

以美国为例的传统体系运作方式是：完成博士后研究后，你需要等待职位空缺，通常找到的教授职位最初也并非终身制。

The way these classical systems work, for instance, in The US, is that you do your postdoc, and then you have to wait to find a position, and usually when you find a professor position, it is not tenured yet.

你成为教授后会有自己的团队和完整配套资源，但仍需等待一段时间才能获得终身教职。

So you become a professor, you have your team, and you have a full package with standings and stuff, but still you need to wait a little bit before being tenured.

法国体系则截然不同且独具特色。

The French system is pretty unique and works exactly the opposite way.

具体而言，高校和研究机构会定期组织竞聘考试，在学者完成博士后阶段后招募研究人员。

That is to say, regularly universities and research institutes are going to organize competitive examinations to recruit researchers once they are done with the postdocs.

只要你在这场竞争中获胜，就能直接获得终身职位——但仅此而已。

And once if you if you succeed in this in this competition, you get tenured, but nothing else.

这就是我目前的职位性质，虽然具体职称有多种，但我的岗位本质上属于chargé de recherche（研究员）。

So this is basically my position, which is called, well, you have different kind of position, but basically the kind of stuff that I have, this is called being a charred rocherche.

这个职位的核心在于：你是公务员身份，相当于终身制的博士后岗位。

The idea is that you are a civil servant and the position is a lifelong postdoc position.

所以实际上，我们获得了终身职位，但没有其他任何附加福利。

So de facto, we are tenured, but without any other perks.

也就是说，我没有自己的团队。

That is to say, don't have my own team.

我不被允许单独指导博士生。

I am not allowed to be the sole supervisor of a PhD student.

我无法申请大型研究经费。

I cannot apply to big grants.

但法国体系是这样运作的：一旦你获得这个仅有终身职位而无其他权限的岗位后，通常不到十年内，你会通过某种超级博士答辩，然后成为正教授。

But the French system works that way, in that once you have this tenured and nothing else position, after a few years, usually less than a decade, you pass some kind of mega PhD defense, and then you become a full professor.

之后你就能拥有自己的团队，申请大型经费这类事情。

And then you're able to have your own team and to have your, to apply for big grants and that kind of stuff.

所以我非常喜欢这种两步走的制度，虽然竞争依然激烈，但至少它能雇佣大量博士后，并确保许多博士后能在学术界找到长期职位，即使他们最终不想成为正教授。

So I really like these two steps systems, which allows people, it's still very competitive, but at least it's allows to to hire a lot of postdocs and to make sure that a lot of postdocs are going to find long term position in academia, even if they don't want to become full professor by the end.

这确实非常独特。

That is very unique.

我从未听说过这种情况。

I never heard about this.

是的。

Yes.

然后逐步地，你可以申请一些小型的资助项目。

And gradually, you can apply for smaller grants.

是的。

Yes.

逐步积累

Building up.

十年后你需要展示什么样的成果？

What is that outcome that you will need to present in ten years?

你提到过，比如那个重要的答辩

You said, like, the big defense.

那具体是什么？

What is that?

为了获得正教授职位，你需要为那次答辩准备些什么？

What do you need to have prepared for that defense to get the full professor position?

所以再次说明，可能我的回答会有点令人失望。

So once again, maybe my answer is going to be a bit disappointing.

我想解释一下，这是我第一次回到法国的学术界，实际上我对法国的学术体系不太熟悉，因为我在法国读了硕士，但在瑞士读了博士，又在美国做了博士后。

I would just like to explain that this is my first time being back in academia in France, and I I'm actually not very familiar with the French academic system because I did my master in France, but then the PhD in Switzerland, and the postdoc in The US.

所以仍有一些环节我还不能完全掌握。

So there are still some elements that I still don't perfectly master.

但基本上，这种超级博士答辩的理念，被称为...

But, basically, the idea of this super PhD defense, which is called an.

好的。

Okay.

其目的是展示自你获得终身教职以来所指导的所有前博士生的研究成果。

The idea is to show all the results of the previous PhD students that you have contributed to supervise since you got tenured.

据我理解，重点固然会放在你自己的成果上，但同时也会关注你的教学工作以及你协助指导的博士生们所取得的成果。

So from what I understood, it is mostly going to be to to focus for sure on your own results, but also on your teaching and on the results that the PhD students that you helped to supervise have been able to obtain.

是的。

Yes.

所以你提到这类似于一种资格认证程序。

So you mentioned it's like a habilitation procedure.

对。

Yes.

我来自立陶宛，我们那里也有资格认证程序。

I'm originally from Lithuania, so we had habilitation procedure.

当然，你先完成博士答辩，然后工作。

Of course, you defense your PhD and then you work.

之后某个阶段你会进行资格认证。

And then at some point you do your habilitation.

我记得在德国也有类似的资格认证，叫什么资格认证博士之类的。

And I think in Germany, is also habilitation, habilitation doctor or something like that.

没错。

Yes.

非常有趣。

Very interesting.

现在，我想稍微深入了解一下你的过去，看看你是如何走到现在这个研究侵入式脑机接口的位置的。

Now, I would like to dive in a little bit into your past and see how did you come to this place where you are working on invasive brain computer interface.

现在转向非侵入式领域，这一切是如何开始的？

Now moving to non invasive, how did this all start?

你是什么时候就已经确定这就是我想做的事情并且正在朝着这个方向努力的？

When did you already know that this is what I want to do and I'm going towards it?

也许再说一次，第三次了，这不是一个很有趣的回答。

Maybe once again, for the third time, not a very interesting answer.

我不知道。

I don't know.

不。

No.

所以我要从我来自的背景来解释这个问题。

So so I I'm going to explain it from from where I'm I'm coming I'm coming from.

我的想法是，我从未在脑海中制定过什么宏伟计划。

The idea is that I never had a grand scheme in mind.

我十年前从未对自己说过，好吧，多少年后我要做这个，然后做那个，接着再做那个。

I never told myself ten years ago, okay, in x years, I'm going to do that, and then I'm going to do that, and then I'm going to do that.

我觉得很多人喜欢这种宏伟的、雄心勃勃的人生规划，这非常好。

I think a lot of people like this kind of grand scheme, very ambitious scheme in mind, which is which is very good.

但大多数时候，职业道路往往取决于各种小意外，所以我发现更有效的方法是进行所谓的'短视优化'。

But most of the time, career path is going to be so dependent on small contingencies that usually I find it more efficient to do what I would call myopic optimization.

就是试着一次只优化一个步骤。

So just try to optimize one step at a time.

好的。

Okay.

我现在处于时间t的这个位置。

I am in this position at time t.

我该采取什么行动，才能在时间t+1时最大化我的幸福感？

Which action should should I take to maximize my happiness just at time t plus one?

我只关注时间t加一时刻，而非展望数年后的未来。

And I'm only going to look at time t plus one and not several years into into the future.

这就是为什么当你查看我的简历时，可能会发现我做过许多不相关的事情，但在每个时间节点t，这都是我当时认为最有趣的选择。

So this is why, when you look at my my CV, I might have done a lot of unrelated stuff, but it's just that at each time step, at each time step t, this is what I felt to be to be the the the most interesting.

于是，我在2018年开始了我的学术之旅。

So, yeah, so I started my academic journey in 2018.

我在瑞士伯尔尼大学Pfister教授的实验室开始了博士生涯，当时那是个完全不研究脑机接口的实验室。

I started my PhD in Switzerland at the University of Bern, in the lab of professor Pfister, at that time in a lab that does absolutely no BCI.

这是一个从事理论神经科学研究的实验室，主要聚焦于大脑和细胞运作的规范性理论，大多时候是开发关于突触如何工作的数学模型，并将其与观察结果进行对比验证。

This was a lab doing theoretical neuroscience, so mostly focusing on normative theories of how the brains and cells are working, mostly developing mathematical models of how synapses are working and confronting them to observations.

我们与不同实验室组成了一个团队，我在理论神经科学实验室，还有计算神经科学团队，以及从事神经形态硬件与计算的团队。

We were in a team with different labs, so I was in the theoretical neuroscience labs, there was a computational neuroscience team, there was a team doing neuromorphic hardware and computations.

所以我们仍在研究神经科学，但团队大多数人拥有物理、数学、工程背景，生物学背景的人很少。

So we are still doing neuroscience, but most of us had backgrounds in physics, mathematics, engineering, very little biology.

是的，当时的想法是，我完成了博士学位，感觉更像是在攻读物理学或统计学的博士学位，但应用到了神经科学领域。实际上Pfister教授是物理学家出身，接受的是理论训练，但非常有意思的是他与许多实验科学家保持着紧密联系。

And yes, so the idea was that, I did my PhD, I had the impression mostly to do a PhD in physics or in statistics, but with application to neuroscience, and actually Professor Pfister is a physicist and a theoretician by training, but what is very interesting is that he has close ties with a lot of experimentalists.

尤其是我的博士联合导师是苏黎世大学的Martin Muller教授，他从事神经信号记录工作。我的博士导师建议我与他共同构建理论，然后通过Martin Muller团队的实验记录和观察数据来验证这些理论。

And especially my PhD co advisor was Professor Martin Muller from the University of Zurich, who is doing recordings, and what my PhD supervisor told me was to build theories with him and then to confront him to see the recordings and observations from the team of Martin Muller.

这非常有意思。

So this was very interesting.

我整个博士阶段都在尝试发展方法论和理论，并将它们与实验观察相对照。

My whole PhD was trying to develop methodologies and theories and to confront them to observations.

我认为自己始终保持着理论学者和计算神经科学家的训练背景，同时持续尝试为具体生物学问题构建量化解决方案。

And I think I kept this idea of being a theoretician and a computational neuroscientist by training, while still trying to build quantitative solutions for concrete biological problems.

这就是我喜欢在学术界做的事情，拥有理论背景，但使其发挥作用并应用于具体的生物学和临床问题。

This is what I like to do in academia, having a theoretical background, but making it useful and applying it to concrete biological and clinical issues.

这就是我博士期间所做的工作。

So this is what we did during my PhD.

基本上，我们开发了优化研究突触方法的算法。

Basically, we came up with algorithms to optimize the way we study synapses.

而这也正是我在匹兹堡做博士后期间所做的，我们运用不同的理论方法来尝试提高PCI控制的精确度。

And hopefully, this is what I did during my postdoc in Pittsburgh, in which we used different theoretical methods to try to improve the accuracy of PCI control.

我能够完成这段博士后研究，要感谢瑞士国家科学基金会的资助。

And I've been able to do this postdoc, thanks to the help of the Swiss National Science Foundation.

他们有一个名为'博士后流动资助'的项目，面向所有在瑞士获得博士学位的人开放，资助他们前往海外进行为期两年的博士后研究并支付薪资。

There is a call for a grant which is called the Postdoc Mobility Grant, which is open to anyone who did their PhD in Switzerland, and which allows them to have their salary paid to do a postdoc for two years abroad.

这是一个非常慷慨的资助计划，我在Officer教授（我的博士导师）和Collinger教授（拟任博士后导师）的共同指导和帮助下成功申请到了该资助，得以在匹兹堡度过了两年时光。

This is a very generous funding scheme, which I applied for, and with the help and the supervision of both Professor Officer or as a PhD supervisor, and Professor Collinger as a would be postdoc supervisor, I've been able to get these grants and to spend two years in Pittsburgh.

最后，在匹兹堡期间我申请了欧洲多个终身教职岗位。

And then finally, while I was in Pittsburgh, I applied to different tenured positions back in Europe.

最终我得以加入法国国家信息与自动化研究所（INRIA），继续从事理论构建工作，并运用计算神经科学来解决临床问题。

I ended up being able to join the INRIA, this Institute for Research in Artificial Intelligence here in France, and to still work on building theories and using computational neuroscience to solve clinical issues.

是的，这太棒了。

Yeah, that is amazing.

非常感谢你指出我们并不总是需要宏大的规划方案。

And thank you so much for pointing that we don't always need a grand schema.

确实如此。

Yes.

为了成功构建我们的职业生涯。

To build our career and build it successfully.

我们可以采取那些循序渐进的步骤。

And we can take those gradual steps.

是的。

Yes.

就像你提到的小型应急方案那样。

Like small contingencies, like you mentioned.

这实际上与你之前在匹兹堡大学所做的研究有相似之处。

And it actually has some parallel to the research you've done, let's say, at the University of Pittsburgh.

没错，那些逐步的改进最终会带来惊人的成果。

Yes, those gradual improvements, but at the end they lead to wonderful results.

我认为这对正在规划职业发展的听众非常重要，因为人们有时会担心自己还没有完整的人生蓝图。

And I think that's very important to know to our listeners who are building their careers, because sometimes people worry, I don't have this grand scheme of my life yet.

也许你并不需要。

And maybe you don't need to.

是的，有些人确实有规划，这很好。

Yes, some people do, and that's okay.

但有些人没有，这也很好，因为职业发展有多种途径。

But some don't, and that's also okay because there are different ways for us to build our careers.

我非常感谢你指出这种方法与科学研究方式之间存在相似之处。

And I really appreciate you saying that there is a parallel between this approach and the way we do science.

其实，我想推荐一本让我深受启发的书。

Actually, would like to to recommend a book that I found to be extremely inspiring.

书名叫《为什么伟大不能被计划》。

It's called Why Greatness Cannot Be Planned.

作者是乔尔·莱曼和肯尼斯·斯坦利，书中精确阐述了科学不在于制定雄心勃勃的项目和宏伟计划，而主要是通过他们所谓的垫脚石来推进工作。

It is by Joel Lehman and Kenneth Stanley, and it's about precisely how science is not about having extremely ambitious project and grand scheme, but mostly work by what they call stepping stones.

所以科学就是一次一小块垫脚石这样逐步推进的。

So science just works one small stepping stones at a time.

他们用来解释这个观点的例子是：如果你穿越回中世纪，召集当时最杰出的头脑，告诉他们拥有无限时间和资源，要在中世纪造出计算机、火箭或核电站。

And the kind of of examples they give to explain that is that if you were to travel back to the middle age and to gather the most brilliant minds of the of the time and to tell them, okay, you have infinite time, infinite resources, build a computer or build a rocket or build a nuclear plant in the middle age.

他们注定会失败。

They're going to fail.

即便拥有无限时间，他们最终仍会失败，因为从零开始建造火箭根本不可能。

Even with infinite times, they're going to fail at some point, simply because it's impossible to say, okay, I'm going to build a rocket from scratch.

从零开始建造核反应堆也是不可能的。

I'm going to build a nuclear pile from scratch.

这只是一系列看似无关的垫脚石，它们可能被随机发现，最终被组合起来创造出伟大的事物——而这些在最初都未被规划。

It's just a succession of seemingly unrelated stepping stones that were discovered maybe randomly, that at some point were brought together to create something great, which was not planned at the beginning.

我喜欢将这种研究方式与我规划职业道路的方式相类比。

And I like to draw a parallel between the way we can consider research that way and the way I try to optimize my different career steps.

你提到现在你在一个专注于AI开发及其医疗应用的研究所工作。

You mentioned that now you work at this institute that focuses on the development of AIs and their application actually in healthcare.

当然，这本身就是一个非常重要且相关的话题。

And of course, that is a whole topic on its own and very relevant.

我想对此进一步探讨一下。

I would like to explore it a little bit more.

首先，根据你的专业见解，你认为在未来十到二十年内，人工智能在临床或医疗保健领域的主要应用方向会有哪些？

First of all, what are in your opinion with your knowledge, what are the main clinical or healthcare applications of AI that we may be able to see in the next, let's say, ten to twenty years?

这个领域的发展趋势是什么？

Where this field is heading?

嗯，我主要会关注这些新的机器学习方法在脑机接口(BCI)领域的应用。

Well, I'm mostly going to focus on the application of these new machine learning methods to to to BCI.

具体来说，利用新型深度神经网络进行运动解码并应用BCI技术，这是我最感兴趣的领域。

Well, the use of new deep neural networks to perform motor decoding and to use BCI, this is something that I am extremely interested in.

我认为直到最近，这个领域还分为两个子方向。

I would say that until recently, the field was separated into two subfields.

一个是机器学习领域，也就是我们所说的人工智能，它提供了许多优秀的解决方案，但这些方案通常相当复杂。

There was a field of machine learning and what we call AI that provides a lot of great solution, which are often pretty complicated.

它们具有非线性特征。

They are nonlinear.

可能需要GPU才能运行。

They might require GPUs to run.

它们涉及数十亿个不同的参数，但从理论上能提供出色的结果。

They involve billions of of different parameters, but they provide great results in theory.

关键在于，直到最近都难以判断这些方法是否真能改善脑机接口中的运动解码效果。

The point is that it's until recently, it was difficult to see whether they are going to indeed improve stuff when it comes to motor decoding in BCIs.

另一方面，你们还有更古老、更简单的线性方法，这些传统上一直用于运动解码。

On the other hand, you have much older linear, simpler methods that have been classically used for motor decoding.

比如我提到的这些简单线性方法——最优线性估计、卡尔曼滤波之类，都是运动解码的经典方法。

So in these simple linear methods, I am mentioning, for instance, optimal linear estimation, Kalman filtering, that kind of stuff, which has been classically used for motor decoding.

但你会问：有没有可能做得更好？毕竟卡尔曼滤波虽好，却是六十年代的技术了。

But the question that you might ask is maybe it's possible to do better because I mean, Kalman filtering is cool, but it was developed in the sixties.

也许是时候该往前迈进了。

Maybe it's time to move a bit further.

所以现状确实令人沮丧：一边是大量炫酷的机器学习方法，但应用到临床时未必有效；另一边是大量久经考验但略显陈旧的传统方法。

So yes, the idea is that the kind of frustrating landscape is that on the one hand, you have plenty of super cool machine learning methods, which may not necessarily work when you apply them to clinical settings, and on the other hand, you have plenty of well proven methods, but which are getting a bit rusty when you compare them to shiny machine learning methods.

最近几年，已有实验室开始尝试应用这些基于机器学习和深度神经网络的方法来改进微运动解码。

And recently, for a few years, some labs have been trying to apply these machine learning and deep neural networks based methods to improve micro decoding.

传统做法是，你将神经活动的记录数据输入解码算法，算法将输出解码器或假肢的速度。

The idea is that classically, you use, you feed the recordings and neural activity into your decoding algorithm, which is going to output the velocity of the decoder or of the prosthetics.

通常，你会使用线性估计或卡尔曼滤波来实现这一点。

And usually, you do that with linear estimation or with Kalman filtering.

那么问题来了，是否有可能用LSTM或卷积神经网络等方法来取代这个卡尔曼滤波器呢？

And the idea would be, will it be possible to replace this Kalman filter with, for instance, an LSTM or a convolutional neural networks?

理论上说，这确实能带来一些改进。

In theory, there are some improvements.

我认为这个领域需要更多证据来证明，从指令滤波转向非线性方法所带来的改进是值得的。

I think the field needs to be a bit more convinced that the improvement is going to be worth the move from command filtering to nonlinear methods.

我认为这确实会带来改进。

I think it's going to indeed bring improvements.

使用非线性机器学习方法最终将改进我们进行脑机接口解码的方式。

The use of nonlinear machine learning methods is ultimately going to improve the way we do BCI decoding.

我只是不确定这是否会是一个巨大的突破，比如性能的爆发式提升，但我认为它仍将是渐进式的改进。

I'm just not sure it's going to be a huge breakthrough, like an explosion of performance, but I think it's still going to be incremental.

这些新方法，感觉它们比卡尔曼滤波和线性方法表现略好一些。

These new methods, it feels like they perform slightly better than Kalman filtering and linear methods.

并非好得离谱，但确实是在朝着正确的方向迈进。

Not insanely better, but still it's going to move into the right direction.

那么你认为是什么阻碍了从那个有点过时、略显笨拙的卡尔曼滤波向这些基于非线性方法的新AI技术过渡呢？

And what do you think still stops this transition from the bot and maybe a little rusty, yes, like Kalman filtering into those new AI based on nonlinear methods?

我不认为

I don't think

本质上存在任何阻碍。

there is anything that blocks per se.

只是特别是在临床环境中工作时，当你已经有了一个运行良好的解决方案时，会存在很大的惯性。

It's just that, especially when you work in a clinical settings, you have a lot of inertia when you have a solution that is already working well.

要转向一个可能表现更好（也可能不会）的新方案需要克服很大惯性，因为唯一能验证效果的方式就是测试，而这需要大量时间，而在与人类参与者合作时时间非常宝贵。

There is a lot of inertia to try to move towards something that might or may not work better, because the only way to to see whether it's going to work better is to test it, which involves a lot of time, and time is very precious when you work with with human participants.

但我认为并不存在任何实质性的障碍。

But I don't think there is any any blockage.

只是我们或许需要进一步加强理论研究者与临床应用人员之间的沟通。

It's just that we probably need to further improve the communication between theory people and clinical and application people.

这正是我所倡导的，也正是我们匹兹堡大学与伯尔尼大学合作项目的核心目标。

This is definitely something that I am advocating for, and this is precisely this project that we have that we have between Pittsburgh and the University of Bern.

匹兹堡大学负责解码研究，拥有庞大的实验设备和出色的人类受试者群体。

The University of Pittsburgh is doing decoding and has a huge setup with amazing human participants.

而在另一端，伯尔尼的研究人员正在开发最先进的机器学习方法。

And on the other end, people in Bern are developing state of the art machine learning methods.

我们当前正在验证这些机器学习方法是否能改进现有的模型解码方式。

And what we are trying to see at the moment is whether this and all these machine learning methods are improving the way we do model decoding.

所以我认为本质上并不存在障碍。

So I don't think there is a blockage per se.

只是需要些时间。

It's just taking some time.

是的。

Yeah.

这没问题。

And it's fine.

在运动控制领域乃至整个BCI领域，你认为当前的主要趋势是什么？这个领域正朝着什么方向发展？

What are the main tendencies in motor And maybe in the whole field of BCI, do you see where the field is heading?

主要的发展方向有哪些？

What are the main directions it's taking?

在你看来，未来五到十年内可能会有哪些新的发展？

And maybe we'll be developing in the, again, next five, ten years, in your opinion, from your point of view?

我会非常谨慎地回答，因为正如我之前提到的，我本身并非百分之百的BCI研究人员。

I'm going to be extremely prudent because as we said about my feedback, I am not per se a 100% BCI researcher.

我来自理论神经科学领域，只是在博士后阶段偶然加入了BCI研究群体。

I'm coming from theoretical neuroscience, and I just happen to have joined the BCI community during my postdoc.

但那些从学术生涯初期就全职投入BCI研究的人，可能会有与我不同的观点。

But probably people who have been full time into BCI since the beginning of their academic journey, they might have different opinions than I do.

以我这个局外人的视角来看，我认为进步主要还是渐进式的。

From my outsider perspective, I would say that the improvements are going to be mostly once again incremental.

回到你之前的问题，在软件方面，我对这些非线性方法仍然抱有很高期望。

On the software side, to go back to your previous question, I'm still very hopeful about these nonlinear methods.

最近的一些论文表明，LSTM、循环神经网络和卷积神经网络确实能够比之前那些经过验证的方法...

There are some recent papers showing that LSTM, recurrent neural networks, convolutional ones are indeed able to improve the performance of decoding as compared to previous and and well approved methods.

理论上，在离线数据上，这种改进相当令人印象深刻。

In theory, on on on offline data, the improvement is pretty impressive.

我们仍需观察这种改进是否足够稳健，并且能否在真人参与的在线实验中同样有效。

We are still to see that this improvement is very robust and is going to be transferable once you try it online with human participants.

但我对此非常乐观，我认为这绝对是当前该领域的发展方向。

But I'm very hopeful, and I think this is definitely the direction that the field is taking at the moment.

在硬件方面，也有一系列持续的改进。

And in terms of hardware, there is a continuous stream of improvements.

比如我们之前提到的电极阵列，它们效果不错，但已经开始显得有些过时了。

For instance, the use arrays that we mentioned previously, they work well, but they are starting to have some gray hair.

具体记不清了

I don't remember exactly, but the patent was from twenty or thirty years ago, because they still work well, but the industry is slowly moving towards improving them.

我特别认为，拥有完全植入皮肤下的设备将会非常有用。

And I especially think that having something that is insanely, that is entirely below the skin is going to be very useful.

例如，神经织网技术前景广阔，因为它不需要任何经皮部件从头骨穿出，这对参与者来说是巨大的负担。

For instance, the neuralling stuff is very promising in that you don't have any percutaneous stuff popping out of the skulls, which is a huge burden for the participants.

因此，我再次强调，我无法预见哪个重大突破会成为脑机接口领域的颠覆性变革。

So once again, I cannot think of one huge breakthrough which is going to be a game changer for the field of BCI.

从我的外部视角来看，进步将会缓慢到来，既来自软件改进——即转向更复杂的解码方案，也来自硬件方面——通过持续优化各种工程方法。

From my outside of view, I think progresses are going to come slowly, both from software improvements, that is to say, moving towards more complicated decoder scheme, and from the hardware side by just just continuously improving the different engineering methods.

是的。

Yes.

谢谢。

Thank you.

你认为想要加入这一领域的人需要具备哪些技能？

And what skills do you think will be required from people who want to join this field?

嗯，可能我有些偏见，因为从专业背景来说，我是工程师，我的博士研究主要集中在理论和计算神经科学领域。

Well, maybe I'm a bit biased because by training, I am an engineer and I mostly did my PhD into theoretical and computational neuroscience.

但我建议人们尽可能多地学习数学、物理和定量分析方面的知识。

But I would advise people to try to do as much math and physics and quantitative stuff as possible.

因为简单来说，如果你从数学密集型ically heavyrically heavy stuffry like math and physics, computer science: "Because simply, if you start studying very math heavy stuff, math, physics, computer science, it's still probably still going to be possible for you to switch to more applied stuff."

Because simply, if you start studying very math heavy stuff, math, physics, computer science, it's still probably still going to be possible for you to switch to more applied stuff.

反过来可能会更困难一些。

The opposite can be a bit more difficult.

也就是说，从生物学和临床学开始，随着时间的推移，要重新回到数学和物理可能会比较困难，这可能会更难。

That is to say, starting with biology and clinics, it can be hard with time to get back to math and to physics, simply because, well, it's probably easier to study math when you are at the beginning of your studies and to then switch to less quantitative fields.

根据经验，我认为反过来会稍微困难一些。

The opposite empirically, I would say is a bit more difficult.

是的。

Yes.

我完全同意你的观点。

I absolutely agree with you.

你认为或看到哪些机会正在向想要进入脑机接口领域的人开放？

And what opportunities do you think or maybe you see are opening for people who want to join the field of brain computer interfaces?

目前我不确定我们之前讨论的这些软硬件小改进是否会增加参与者数量或实验室数量，这是我期望看到的。

I am not sure at the moment what these little software and hardware improvements that we discussed previously, whether they are going to increase participants, whether they are going to increase the number of labs, this is something that I hope for.

目前从事脑机接口研究的实验室和参与者数量都非常少。

For the moment, the number of labs that do BCIs and the number of participants is very small.

参与者就像是宇航员一样稀少。

It's like the the participants are like astronauts.

只有几百人上过太空，也只有几百人使用过脑机接口设备。

They're just a few hundreds of people who have been into space, and there are just a few hundreds of people who have been using BCIs.

所以我真正期待的是——虽然不确定能否实现——希望有更多实验室开放，招募更多参与者。只有这样才能普及这些方法，让它们不仅限于科研环境中的少数受试者，而是能更广泛地应用于临床领域。

So what I really hope for, I'm not sure whether it's going to take place, but I what I really hope for is for more labs to open, more participants to be recruited, this is simply how we are going to democratize these methods and to make them technologies that are useful not only for just some participants in scientific settings, but which are going to be used more widely in the clinical settings.

我不确定这能否很快实现，但这绝对是我会大力倡导的方向。

I'm not sure this is going to happen anytime soon, but this is definitely something that I would advocate for.

或许您了解法国在支持神经科技或脑机接口项目方面的现状？如果不清楚也完全理解。

And maybe you are aware, if not, of course, I understand if you are aware of the situation in France in terms of supporting neurotech or BCI based projects.

随着技术发展，目前的资助支持和资金投入是保持稳定、正在减少，还是有所增加呢？

Is the grant support, financial support is being stable, decreasing, or maybe increasing as technologies are developing?

我想说，与美国相比，我们在欧洲面临的主要困难是进行侵入性PCI的可能性或缺乏。

I would say the main difficulty that we have in Europe as compared to The US is the possibility or the lack thereof to do invasive PCIs.

非侵入性PCI很酷，但我的意思是，任何PCI解决方案都有其优缺点。

Non invasive PCIs is cool, but I mean, any PCI solution has its pros and its cons.

通过皮层内记录，你可以获取单个神经元的活动，并拥有海量信息进行精细解码。

With intracortical recordings, you have the activity of individual neurons and you have an insane amount of information to do fine and decoding.

但本质上这是皮层内操作，许多参与者可能不同意将电极植入他们的头骨。

But per se, this is intracortical, and a lot of participants might not agree to have some electrodes burnt into their skull.

另一方面，欧洲和法国的大多数脑机接口实验室都专注于脑电图技术，这非常实用，特别是因为对参与者来说实施起来要容易得多，但你无法获得相同的信息解码质量和自由度。

On the other hand, most BCI labs in Europe and in France are focusing on EEG, which is very practical, especially because this is way, way easier to use for the to implement for the participants, but you don't have the same degree of quality and the same number of degrees of freedom that you can decode information from.

我认为目前欧洲与美国的主要区别在于，在欧洲很难获得进行皮层内侵入式脑机接口研究的授权。

I would say the main difference between Europe and The US at the moment is the fact that it's very difficult to obtain authorization to do intracortical invasive PCI in Europe.

我感觉这种情况正在改变，特别是法国的格勒诺布尔大学正在开发许多皮层脑电图解决方案，我希望这些解决方案能越来越普及。

I have the impression that this is changing, especially about the University of Grenoble in France, which is developing a lot of ECOG solutions, and I hope that these solutions are going to be more and more widespread.

再次强调，我不确定这是否会发生，也不确定其发展速度，但这确实是我们目前所倡导的方向。

Once again, I'm not sure this is going to take place, and I'm not sure at what pace this is going to take place, but this is definitely something that we are advocating from at the moment.

我所在的机构有一个名为PIQ的资助计划。

There is the institute I am in as a funding scheme, which is called the PIQ.

不记得具体代表什么，但其理念是资助高风险项目。

Don't remember what it stands from, but the idea is that it's a funding scheme that funds very high risk project.

传统上，资助项目通常授予——我不会说是无聊的项目——但风险不太高的那些。

So classically, and grants, they are awarded to, I will not say boring project, but not too risky one.

如果你带着疯狂的想法和高风险去找资助机构，确实不太可能获得资金支持。

If you come to a funding agency with a crazy idea, with a lot of risk, yeah, you're not going to get your your fundings.

而这个资助计划恰恰针对高风险但可能带来高回报的项目。

Precisely, this funding scheme is aimed at very high risk, but high possible reward project.

目前我们正与一些同事讨论申请该计划用于BCICT项目的可能性。

And with some colleagues, we are currently discussing the possibility to apply to one precisely for a BCICT.

未完待续，但这类资助高风险项目的举措确实极具前景，我积极倡导此事，我们正在探讨利用该计划在欧洲开发更具侵入性的脑机接口记录技术。

To be continued, but this kind of initiative, that is to say, funding high risk projects, this is definitely something that is really promising, that I am advocating for, and we are discussing the possibility to use one developing more invasive BCI recordings in Europe.

是的。

Yes.

希望几年后我们能再见面，也许到时你可以更新一下你的新项目进展以及这一切的结果如何。

I hope that we can meet in a couple of years and maybe you can give an update of already your new projects and how did this all turned out.

关于BCI奖项，我想为我们的听众提几个问题，特别是那些计划申请的人。

And just few questions about the BCI award for our listeners, for those who are planning to apply.

你当初申请BCI奖项的具体原因是什么？

What was the reason for you to actually apply for the BCI award?

为什么决定要这么做呢？

Why did you decide to do that?

嗯，正如我之前提到的，在加入匹兹堡大学实验室之前，我并未融入BCI社群。

Well, yeah, as we discussed before joining the lab at the University of Pittsburgh, I was not into the BCI community.

所以我完全是个局外人。

So I I was completely an outsider.

我并不是说BCI社群是个小圈子，但它确实是个成熟的社群，有自己的知名人物、研究方法和会议体系。

And I'm not saying that the BCI community is a is a clique, but it's a well established community with with its celebrities, its methods, its conferences.

因此作为一个外来者加入这个BCI社群，对我来说还是有些令人却步的。

So it was a bit intimidating for me to join this BCI community from the from the outside.

我记得在SFN的一次会议上，应该是在华盛顿，有人找Jen Kolinger要自拍或签名。

And I remember at one conference at SFN, I think in Washington, there were some people asking Jen Kolinger for for for selfies or for autograph.

我当时就想，哇，我真的是在和重要人物共事。

And I was like, wow, I'm really working with with important people.

我真的是在这里和大人物一起工作，这让我感到有些害怕和紧张。

I'm really working with bigwigs in here, which made it a bit scary and intimidating for me.

我曾与BCI社区的许多知名人士进行过Zoom会议。

I had Zoom meetings with plenty of celebrities of the BCI community.

我认为申请这个奖项，对我来说是直面自己想法并向社区展示的一种方式。

And I think that applying to this award, it was a way for me to confront my ideas and to be able to show them to the community.

我觉得效果不错。

And I think it worked well.

效果很好，这得益于BCI奖项的展示和宣布。

It worked well because of the presentations and the announcement of the BCI awards.

它正好在上届SFM会议期间举行。

It took place exactly during the last SFM conference.

当时我正在酒店房间里观看颁奖直播，结束后我立即前往会议中心，有幸与其他获奖者交流——他们其实是我经常在Zoom会议上见到的芝加哥大学的研究人员，但之前从未有机会线下见面。

And so I was in my hotel room during the presentation, and right after I went to the conference center and I had the opportunity to discuss with the other people who got awarded, which are actually people from the University of Chicago that I had Zoom meeting with regularly, but I never had the opportunity to in person exchange with them.

这感觉真的很棒。

So it was really cool.

是的。

Yes.

是的。

Yes.

感谢你提到这一点，因为你的意思是说，即使刚进入这个领域的新人也可以申请这个奖项。

Thank you for mentioning that because what you are saying is if you are new to the field, you can apply for the award.

这是其中一种方式。

This is one of the ways.

没错，虽然不是唯一途径，但确实是融入社群、建立人脉、结识同行的方式之一。

Yes, not all, but one of the ways how to connect with the community, how to network, how to get to know people.

这确实是个绝佳的社交工具。

And again, that's a great instrument for doing that.

你对如何让申请成功有什么建议吗？

Any suggestions do you have for people on how to make their submission successful?

因为你显然成功了。

Because you're obviously was.

是吗？

Yes?

我想说的是不要过分执着于奖项。

I would say try not to be obsessed with awards.

这可能听起来有点反直觉，但尽量不要过分痴迷奖项。

That might be that might be counterintuitive, but try not to be too obsessed with awards.

总的来说，在科研领域会遇到很多挫折。

In general, in science, there is a lot of frustration.

你会申请很多东西：很多职位、很多奖项、很多会议、很多期刊。

You are going to apply to a lot of stuff, to a lot of position, to a lot of awards, to a lot of conference, to a lot of journals.

99%的情况下你会被拒绝，得不到回复，或者收到否定答复，这很正常。

And 99% of the time you are going to get rejected, you are not going to get an answer, you are going to be answered no, and it's fine.

这就是科学研究的常态。

This is all science works.

真正重要的不是那99%的拒绝。

What matters is not the 99% of the rejections.

这些拒绝并不意味着你的科研做得不好。

They don't mean that you are doing poor science.

它们也不代表你不是一位优秀的科学家。

They don't mean that you are not a good scientist.

你需要关注的是那1%的成功机会。

What you need to focus on is this 1%.

仅仅因为竞争非常激烈，所以你不应该因为论文被拒、奖项落选或类似的事情而灰心丧气。

Simply because this is very competitive, so you should not be let down by papers being rejected by an award that you missed or all that kind of stuff.

至于申请本身，我认为最好的方法——无论是申请BCI奖项、职位、新项目资助还是投稿论文——就是尽可能让更多人审阅你的材料，因为他们能告诉你内容是否清晰、创新点是否突出。

As for the application itself, I would say the best way, and this is true for an application to the BCI award, for an application to a position, for a new grant submission, for a new paper, have as many people as possible review it, because they are going to be able to tell you whether it's clear, whether the novelty is well highlighted.

因此无论你需要提交什么材料，无论是BCI奖项还是其他申请，都要相信同行并听取他们的建议。

So whatever you need to submit something, whether this is for the BCI award or for something else, just trust your peers and have them give you suggestions.

判断某些同行无法提供建设性反馈也是可以的。

It's also fine to judge that some of your peers are not going to give you constructive feedbacks.

我认为这正是你作为科学家成长的方式。

And I think this is the way you grow as a scientist.

最初，你会全盘接受所有反馈。

At first, you are going to take everything that is sent at you.

你会把这些反馈当作针对个人。

You're going to take it personally.

但随着成长，你会开始思考：也许这个反馈并不具有建设性。

And then as you grow up, you're going to think, oh, maybe this feedback wasn't very constructive.

选择忽略某些反馈也是完全可以的。

Maybe it's fine to just ignore it, which is fine.

所以，无论你提交什么，都要尽量获取同行的反馈意见。

So, yeah, whatever you try to submit, try to have the feedbacks of your peers.

很好的建议。

Great suggestion.

还有一个问题，我们的播客叫做《实现不可能》。

And another question is our podcast is called Doing the Impossible.

在你生活或职业生涯中，有没有什么事情你曾认为不可能，但最终却成功实现并使之成为可能的？

Was there anything in your life or in your career that you think of as impossible, but you actually made it work, made it possible?

你是如何做到的？

And how did you do that?

这与PCI关系不大，但我还是想分享一下。

It's very slightly PCI related, but I'm going to give it a shot.

正如我之前所说，在进入学术界和开始攻读博士之前，我接受过工程学培训，尤其是航天工程领域。

As I said before joining academia and before starting my PhD, I had a training in engineering and especially in space engineering.

在成为脑科学专家之前，我其实是个太空迷。

Before being a brain nerd, I was a space nerd.

我在法国获得了工程学硕士学位，专攻航天工程。

So I got my master's in engineering, in space engineering in France.

在进入学术界之前，我曾在法国空客工业部门工作过两年。

And before joining academia, I've been working for two years in the industry with Airbus in France.

我一直以来的梦想就是成为一名宇航员。

And what I've always dreamt of was to become an astronaut.

非常幸运的是，我曾在2016年加拿大和2022年欧洲的宇航员选拔中两次成功落选。

Very luckily, I've been able to successfully failed twice for an astronaut selection, the 2016 Canadian one and the 2022 European ones.

但我并没有气馁。

I'm not deterred.

我会继续尝试下一次选拔。

I'm going to try for the next ones.

但还有一件非常酷的事就是参加抛物线飞行。

But something that is very cool also is to do some parabolic flights.

我不知道你是否听说过这些抛物线飞行，它们在大飞机上进行，沿着特定轨迹飞行以获得几秒钟的微重力环境，用于微重力科学研究。

So I don't know if you if you've heard these these parabolic flights taking place in big aligners who are following these kind of trajectories in order to have a few seconds of microgravity and to do microgravity science.

就在几周前，我被选中作为其中一次飞行的参与者，非常有趣的是，飞机上有个团队正在进行脑电图实验，研究大脑活动——特别是脑电图活动——如何受微重力和超重力影响。

And so I've been selected as a participant for one of these slides just a few weeks ago, and very interestingly, there was a team in the plane doing an EEG experiment, just analyzing how your brain and your brain activity and especially your EEG activity is impacted by microgravity or hypergravity.

特别有意思的是，这些设备正是由组织BCIOI的那家公司借出的。

And very interestingly, the the equipment was lent by the very same company that is organizing the the the BCIOI.

嗯。

Yeah.

GTEC。

GTEC.

是的。

Yes.

Gtech公司。

Gtech company.

对。

Yeah.

一切都是相关联的。

Everything is related.

没错。

Exactly.

对我来说看到飞机上的Gtech贴纸很有趣。

It was funny for me to see the the the the Gtech sticker and the plane.

我当时就想，我认识这些人。

Was like, I know these guys.

是的。

Yes.

这简直太棒了。

That is absolutely amazing.

是什么让你投身航天事业？

What calls you into space?

啊，因为它很酷。

Ah, because it's cool.

不，我想说，是的，我想说是太空研究。

No, I I would say, yeah, I would say space research.

这仍然是应用理论并尝试构建完整问题的一种方式。

This is still a way to apply theories and to try to build complete complete problems.

许多为太空开发的技术后来在地球上发现了完整的应用，这正是我想为之贡献的领域。

Plenty of technologies that were developed for space then found out to us some complete applications on Earth, and this is something I would like to contribute to.

是的。

Yes.

这太棒了。

That is amazing.

当然，我祝愿你在太空探索和脑机接口研究的旅程中一切顺利，你正在做非常了不起的事情。

And, of course, I wish you all the best and in your space journey and in your BCI journey, you are doing amazing things.

我认为这次播客中非常重要的启示是，感谢你的分享，事情可以逐步推进，但它们确实能产生成果。

And I think the very important note from this podcast, and thank you for that, that things can be done gradually, but they really create results.

所以谢谢你。

So thank you.

不客气。

Take your time.

专注于做好科学研究。

Focus on doing good science.

记住，我们非常幸运能够获得报酬来进行科学研究。

Never forget that we are insanely lucky to be paid to do science.

永远别忘了享受科研的乐趣。

Never forget to to have fun doing science.

慢慢来，别着急。

Just take your time.

坚守道德准则，保持优质成果，成功自然会随之而来。

Keep good ethics, keep good results, and it's going to come along.

非常感谢。

Thank you so much.

卡米尔，听众们了解你工作和联系你的最佳方式是什么？

Camille, what is the best way for our listeners to learn about the work you do and connect if possible?

最好的方式可能是定期查看我们研究所的官网，INRIA。

Probably the best way is to check regularly the website of the institute is called INRIA.

网站上会讨论我们正在进行的各项改进。

There is a website discussing the different improvement that we are doing.

我们的实验室目前仍在筹建阶段。

Our labs is still in the process of its creation.

所以未来可能会有很多新消息陆续发布。

So probably plenty of news are going to pop up recently in the future.

是的。

Yes.

或许大家也可以通过LinkedIn联系你。

And probably people can connect with you on LinkedIn.

没错，你确实在用LinkedIn。

That yes, you're on LinkedIn.

对。

Yes.

当然可以。

Of course.

没问题。

Of course.

太好了。

Wonderful.

非常好。

Very good.

所以我会问你不少问题。

So I'll ask you quite a few questions.

有没有什么我可能没问到，但你还想和听众分享的事情？

Is there anything that maybe I didn't ask you, but you still would like to share with our listeners?

我想没有。

I don't think so.

我想没有。

I don't think so.

是的。

Yeah.

你你问过这个问题，关于我在科学领域有哪些具体人物作为灵感来源。

You you asked this question about what what specific figures I have in mind as an inspiration in science.

对。

Yes.

我很抱歉。

I'm sorry.

我想我没有。

I don't think I have any.

有很多人让我觉得非常令人钦佩，但我觉得说'嘿'会很不公平。

There are a lot of people that I really that I found really impressive, but I think it would be very unfair for me to say, hey.

我更喜欢居里夫人胜过爱因斯坦。

I really prefer Marie Curie over Albert Einstein.

我认为这样说不公平。

I don't think that would be fair of me.

我们都是站在巨人的肩膀上，我不想只挑选其中一位。

We are all sitting on the shoulders of giants, and I don't want to pick just one single one.

是的。

Yes.

这也是个非常美好的回答。

And that's a beautiful response as well.

谢谢。

Thank you.

谢谢你，卡米尔。

Thank you, Camille.

采访你真是我的荣幸。

It was a true pleasure to interview you.

亲爱的《神经载体》播客听众们，感谢你们加入我和这些出色的嘉宾，共同探索神经科学与神经技术领域的职业之旅。

Dear Neurocarriers podcast listeners, thank you for joining me and my incredible guests on this exciting journey into careers in neuroscience and neurotechnologies.

希望这些将突破性想法转化为实际影响的故事能激励你们。

I hope you've been inspired by the stories of those turning groundbreaking ideas into impactful realities.

若您需要更多关于神经领域职业发展的指导，可以预约与我——K博士（你们的播客主持人兼神经职业教练）的免费咨询，我们所在的神经方法研究所是唯一专注于神经科技专业人士需求的职业服务机构。

If you are looking for more guidance on advancing your neuro career, book a free consultation with me, Doctor K, your podcast host and neuro careers coach at the Institute of Neuro approaches, the only career service dedicated specifically to the needs of professionals in Neurotech.

让我们携手迈向你在神经科技领域的成功下一步。

Let's take the next step toward your neurocareer success together.

除了免费咨询外，神经方法研究所还提供多种服务，旨在帮助你在神经科技领域蓬勃发展。

In addition to free consultations, the Institute of Neuro approaches offers a variety of services designed to help you thrive in neurotech.

第一，职业发展规划。

First, professional development planning.

我们帮助您制定个性化的职业发展计划，找出知识短板，并为您提供在神经科学与神经技术领域实现目标所需的工具。

We help you create a tailored career development plan, identify knowledge gaps, and equip you with the tools needed to achieve your goals in neuroscience and neurotechnologies.

第二，简历和求职信审核。

Second, a resume and cover letter review.

获取专家反馈，学习如何制作专为神经科技领域职位机会量身定制的申请文书。

Get expert feedback on how to craft documents that are specifically designed for job opportunities in the neurotech field.

第三，面试准备。

Third, interview preparation.

通过模拟面试和针对性反馈来提升面试技巧，包括软技能、神经科技硬技能（脑机接口、临床脑电图、医疗器械方向）、现场编程及演讲准备等。

Sharpen your interview skills through mock interviews and targeted feedback including soft skills, hard skills in neurotech, bci, ceg, med, live coding and presentation preparation.

第四，

Four.

人脉拓展与求职策略。

Networking and job search strategy.

学习如何在竞争激烈的神经科技就业市场中有效建立人脉并发现工作机会，让自己脱颖而出。

Learn how to effectively network and uncover job opportunities while standing out in the competitive neurotech job market.

谈判与薪资建议。

Negotiation and salary advice.

获取专家指导，学习如何协商薪资福利，并确定您目标职位的合理薪资范围。

Receive expert guidance on negotiating salaries and benefits and determining a fair salary range for your desired position.

以上所有内容及更多信息均可访问www.neuroapproaches.org获取。

All this and more can be found at www.neuroapproaches.org.

再次强调，请访问www.neuroapproaches.org。

Again, www.neuroapproaches.org.

那您还在等什么呢？

So what are you waiting for?

让我们共同探索新兴职业领域的成功之路，将不可能变为可能。

Let's navigate the path to success in the world of new careers together and turn the impossible into possible.

我们期待与您合作。

We're looking forward to working with you.