Sheila Nirenberg: A prosthetic eye to treat blindness

譯者: Ana Choi 審譯者: Hsin Haydn

I study how the brain processes

我研究大腦如何

information. That is, how it takes

處理信息。也就是說，

information in from the outside world, and

它如何從外界採取信息，

converts it into patterns of electrical activity,

並將之變換成電能動性的模式，

and then how it uses those patterns

然後它是如何使用這些模式

to allow you to do things --

讓你做種種事情 --

to see, hear, to reach for an object.

看得到，聽得到，伸出以觸摸到物體。

So I'm really a basic scientist, not

所以我只是一個基本的科學家，

a clinician, but in the last year and a half

而不是一個發聲學家，但在過去一年半中，

I've started to switch over, to use what

我已經開始發生變化，以使用着

we've been learning about these patterns

我們一直在學習的活動模式，

of activity to develop prosthetic devices,

來發展義眼裝置，

and what I wanted to do today is show you

而我今天想要做的便是告訴你

an example of this.

這樣的一個例子。

It's really our first foray into this.

這真是我們首次涉足這個。

It's the development of a prosthetic device

它是一項為治療失明的

for treating blindness.

義眼裝置的發展。

So let me start in on that problem.

讓我開始談談這個問題。

There are 10 million people in the U.S.

在美國有10萬

and many more worldwide who are blind

和在全球有更多的人是盲人，

or are facing blindness due to diseases

或由於視網膜疾病，

of the retina, diseases like

黃斑變性等疾病

macular degeneration, and there's little

正面臨失明，可以為幫助他們

that can be done for them.

做的事情並不多。

There are some drug treatments, but

是有一些藥物治療，

they're only effective on a small fraction

但只在一小部分人的有效。

of the population. And so, for the vast

因此大多數患者，

majority of patients, their best hope for

他們最有希望重新恢復視力

regaining sight is through prosthetic devices.

便是通過義眼的裝置。

The problem is that current prosthetics

問題是, 目前的義眼

don't work very well. They're still very

的運作不是很好。它們仍然

limited in the vision that they can provide.

只可以提供非常有限的視野。

And so, you know, for example, with these

所以你知道，例如，有了

devices, patients can see simple things

這些設備，患者可以看到基本的東西

like bright lights and high contrast edges,

如明亮的燈光和高對比度的邊緣，

not very much more, so nothing close

幾乎是此已以，一直沒有

to normal vision has been possible.

接近正常視力的可能。

So what I'm going to tell you about today

那麼今天我要告訴你的

is a device that we've been working on

是一個我們一直在研製的設備，

that I think has the potential to make

我認為它有可能大大有所作為，

a difference, to be much more effective,

更要有效得多，

and what I wanted to do is show you

我想做的是要向你展示

how it works. Okay, so let me back up a

它是如何運作的。好，第一，讓我返回

little bit and show you how a normal retina

一點，並且告訴你正常的視網膜

works first so you can see the problem

是如何運作，是以你可以看到

that we were trying to solve.

我們要試圖解決的問題。

Here you have a retina.

這裡是一塊視網膜。

So you have an image, a retina, and a brain.

有一個圖像，一塊視網膜和一個腦。

So when you look at something, like this image

當你看着東西，像這樣的

of this baby's face, it goes into your eye

嬰兒面孔的影像，它進入你眼睛,

and it lands on your retina, on the front-end

到達你的視網膜上的前端細胞，

cells here, the photoreceptors.

這裡的受感光器。

Then what happens is the retinal circuitry,

然後將會發生的是視網膜電路，

the middle part, goes to work on it,

中間的部分，會去勞動，

and what it does is it performs operations

執行着它的操作工作，

on it, it extracts information from it, and it

從中提取信息，

converts that information into a code.

並將之轉換成代碼的信息。

And the code is in the form of these patterns

而代碼是在以這些模式的

of electrical pulses that get sent

電脈被發送到大腦，

up to the brain, and so the key thing is

所以關鍵是該圖像

that the image ultimately gets converted

最終要被轉換成代碼的形式。

into a code. And when I say code,

而當我說代碼，

I do literally mean code.

我意思確實是指代碼。

Like this pattern of pulses here actually means "baby's face,"

例如這種模式的脈搏在這兒實際上表示「寶寶的臉」，

and so when the brain gets this pattern

所以當大腦得到這種模式的脈搏時，

of pulses, it knows that what was out there

它便知道在這裡

was a baby's face, and if it

是一個嬰兒的臉，而如果它

got a different pattern it would know

得到了不同的模式，它就知道，

that what was out there was, say, a dog,

在這裡是，比方說，一隻狗，

or another pattern would be a house.

或另一種模式將是一所房子。

Anyway, you get the idea.

總之，你明白這是怎麼一回事。

And, of course, in real life, it's all dynamic,

當然，在現實生活中，它是活躍的，

meaning that it's changing all the time,

它每時每刻都在變化，

so the patterns of pulses are changing

所以模式的脈搏每時每刻也在改變，

all the time because the world you're

因為你看到的世界

looking at is changing all the time too.

每時每刻都在變化。

So, you know, it's sort of a complicated

可想而知道，這是一個複雜的事情。

thing. You have these patterns of pulses

你的眼睛每毫秒

coming out of your eye every millisecond

會發出這些脈搏的模式

telling your brain what it is that you're seeing.

告訴你的大腦你看到的是什麼。

So what happens when a person

當一個人的視網膜

gets a retinal degenerative disease like

得到像黃斑變性的退行性疾病

macular degeneration? What happens is

會怎麼？情況會是，

is that, the front-end cells die,

前端細胞死亡，

the photoreceptors die, and over time,

光感接收器死亡，隨著時間流逝，

all the cells and the circuits that are

所有連接到它們的

connected to them, they die too.

細胞和線路，它們也會死掉。

Until the only things that you have left

直到你剩下唯一的這些細胞,

are these cells here, the output cells,

這些輸出的細胞，

the ones that send the signals to the brain,

負責發送信號到大腦的細胞，

but because of all that degeneration

但因為所有的退化

they aren't sending any signals anymore.

使得它們沒有再發出任何信號。

They aren't getting any input, so

它們沒有得到任何輸入的信號，

the person's brain no longer gets

所以大腦不再得到

any visual information --

任何視覺的信息 --

that is, he or she is blind.

也就是說，他或她是盲的。

So, a solution to the problem, then,

因此，問題的解決方法便是，

would be to build a device that could mimic

要建立一個可以模仿

the actions of that front-end circuitry

前端線路的動作和發送信號

and send signals to the retina's output cells,

到視網膜的輸出細胞的裝置，

and they can go back to doing their

以至它們可以重回

normal job of sending signals to the brain.

做其正常的工作--發送信號到大腦。

So this is what we've been working on,

這便是我們一直在做的工作，

and this is what our prosthetic does.

我們的義眼便是會這樣做。

So it consists of two parts, what we call

它由兩部分組成，我們稱

an encoder and a transducer.

一個為編碼器和一個傳感器。

And so the encoder does just

那編碼器就正如我說的：

what I was saying: it mimics the actions

它會模擬

of the front-end circuitry -- so it takes images

前端線路的作用 -- 將圖像

in and converts them into the retina's code.

轉換成視網膜的代碼。

And then the transducer then makes the

然後那傳感器會製造輸出

output cells send the code on up

代碼的細胞發送到大腦，

to the brain, and the result is

結果便是一個義體的視網膜

a retinal prosthetic that can produce

可以產生正常的

normal retinal output.

視網膜輸出。

So a completely blind retina,

因此，一個完全失明的視網膜，

even one with no front-end circuitry at all,

甚至沒有任何前端線路，

no photoreceptors,

沒有感光器，

can now send out normal signals,

現在便可以發送出正常的信號，

signals that the brain can understand.

大腦可以理解的信號。

So no other device has been able

沒有其他的設備已經

to do this.

能夠做到這一點。

Okay, so I just want to take

好, 我只是想

a sentence or two to say something about

用一兩句話來說說

the encoder and what it's doing, because

編碼器和它在做什麼，因為

it's really the key part and it's

它是真正的關鍵部分，

sort of interesting and kind of cool.

它很有趣和酷。

I'm not sure "cool" is really the right word, but

我不知道「酷」是否正確的用詞，

you know what I mean.

但你明白我的意思。

So what it's doing is, it's replacing

它在做什麼呢，

the retinal circuitry, really the guts of

它在取代視網膜線路，真正

the retinal circuitry, with a set of equations,

視網膜線路的腸道，以一個方程組，

a set of equations that we can implement

一個我們可以實現在一塊芯片上

on a chip. So it's just math.

的方程組。那只不過是數學。

In other words, we're not literally replacing

換句話說，我們不是取代

the components of the retina.

視網膜的組成部分。

It's not like we're making a little mini-device

我們不是像為每個不同類型的細胞

for each of the different cell types.

正在製造一個小小的設備。

We've just abstracted what the

我們只不過是摘要

retina's doing with a set of equations.

視網膜在做什麼的一組方程。

And so, in a way, the equations are serving

因此，在某種程度上，方程是作為

as sort of a codebook. An image comes in,

碼本的一個排序。一個圖像進來，

goes through the set of equations,

會通過方程組，

and out comes streams of electrical pulses,

走出來的是電脈衝流，

just like a normal retina would produce.

就像一個正常視網膜會產生的。

Now let me put my money

現在讓我實際的

where my mouth is and show you that

演示給你看，我們事實上

we can actually produce normal output,

是可以製造正常的輸出，

and what the implications of this are.

而且這會有什麼影響。

Here are three sets of

這裡有三種

firing patterns. The top one is from

發射的模式。在最上面是

a normal animal, the middle one is from

從一隻正常的動物，中間的是從

a blind animal that's been treated with

一隻被此編碼器傳感設備處理過的

this encoder-transducer device, and the

盲的動物，而最下面的是

bottom one is from a blind animal treated

一隻被設置標準義眼的

with a standard prosthetic.

盲的動物。

So the bottom one is the state-of-the-art

最下面的是外面可得到的

device that's out there right now, which is

最新水平的設備，基本上

basically made up of light detectors,

是由光探測器製作，

but no encoder. So what we did was we

但沒有編碼器。我們所做的

presented movies of everyday things --

就是我們放映日常瑣事的短片--

people, babies, park benches,

人，嬰兒，公園長椅，

you know, regular things happening -- and

你所知道的，經常發生 的事情--

we recorded the responses from the retinas

我們記錄了這三組動物的

of these three groups of animals.

視網膜的反應。

Now just to orient you, each box is showing

現在只是在引導你，每個方塊呈現出

the firing patterns of several cells,

幾個細胞的放電模式，

and just as in the previous slides,

就像在之前的幻燈片，

each row is a different cell,

每一行是一個不同的細胞，

and I just made the pulses a little bit smaller

而我只是將脈搏弄小一點，

and thinner so I could show you

薄一點，以便我可以演示給你

a long stretch of data.

看到一串長長的數據。

So as you can see, the firing patterns

正如你可以看到，從有編碼器

from the blind animal treated with

傳感處理的

the encoder-transducer really do very

盲的動物的發射模式與

closely match the normal firing patterns --

正常發射模式確實非常密切的相稱 --

and it's not perfect, but it's pretty good --

它並不是完美，但實在是相當不錯 --

and the blind animal treated with

看看被設置標準義眼的

the standard prosthetic,

盲的動物，

the responses really don't.

反應真的不怎樣發射。

And so with the standard method,

標準的方法，

the cells do fire, they just don't fire

細胞是有發射，但它們不發射出

in the normal firing patterns because

正常的模式, 因為它們到從發射模式的反應呢？

they don't have the right code.

沒有正確的代碼。

How important is this?

這是有多重要呢？

What's the potential impact

對一個病人視力的能力

on a patient's ability to see?

有什麼潛在影響呢？

So I'm just going to show you one

所以我要給你看看

bottom-line experiment that answers this,

一個回答了這個問題的底線實驗，

and of course I've got a lot of other data,

當然我還有很多其他的數據，

so if you're interested I'm happy

如果你有興趣我會很高興地

to show more. So the experiment

顯示更多。這個實驗被稱為

is called a reconstruction experiment.

一個重建實驗。

So what we did is we took a moment

我們做的是從這些錄像

in time from these recordings and asked,

抽出一刻然後問,

what was the retina seeing at that moment?

究竟那一刻視網膜看到的是什麼？

Can we reconstruct what the retina

我們可否重建視網膜

was seeing from the responses

看到的是從

from the firing patterns?

發射模式的反應呢？

So, when we did this for responses

因此，當我們便實驗在

from the standard method and from

用標準的方法,

our encoder and transducer.

和用我們的編碼器和傳感器。

So let me show you, and I'm going to

讓我給你看看，我要首先給你

start with the standard method first.

看標準的方法。

So you can see that it's pretty limited,

你可以看到它是相當有限，

and because the firing patterns aren't

因為發射模式不是

in the right code, they're very limited in

正確的代碼，它們可以告訴你的

what they can tell you about

東西是十分有限。

what's out there. So you can see that

你可以看到，

there's something there, but it's not so clear

是有東西有，但那東西是什麼

what that something is, and this just sort of

不是非常清楚，這正如

circles back to what I was saying in the

我在開始時所說的,

beginning, that with the standard method,

用標準的方法，

patients can see high-contrast edges, they

病人可以看到高對比度的邊緣，

can see light, but it doesn't easily go

但頂多就是這樣。

further than that. So what was

那麼究竟是什麼形象？

the image? It was a baby's face.

這是一個嬰兒的面孔。

So what about with our approach,

那, 用我們的方法，

adding the code? And you can see

添加代碼, 會如何？你可以看到

that it's much better. Not only can you

它是好得多。你不僅可以看到

tell that it's a baby's face, but you can

這是一個嬰兒的面孔，你可以看到，

tell that it's this baby's face, which is a

是這一個嬰兒的臉孔，這是一個

really challenging task.

非常具有挑戰性的任務。

So on the left is the encoder

在左邊的是單獨的編碼器，

alone, and on the right is from an actual

在右邊是從實際的

blind retina, so the encoder and the transducer.

盲的視網膜，使用編碼器和傳感器。

But the key one really is the encoder alone,

但真正的重點是單獨在編碼器，

because we can team up the encoder with

因為我們可以與不同的

the different transducer.

傳感器聯合組隊。

This is just actually the first one that we tried.

這只是我們試用的第一個。

I just wanted to say something about the standard method.

我只想說一些有關標準的方法。

When this first came out, it was just a really

當這個結果出來時，它只是一個

exciting thing, the idea that you

令人非常興奮的事情，一個使

even make a blind retina respond at all.

盲人視網膜有回應的概念。

But there was this limiting factor,

但當時是有限制因素，

the issue of the code, and how to make

代碼的問題，以及如何

the cells respond better,

使細胞有更好地反應，

produce normal responses,

產生出正常的反應，等等，

and so this was our contribution.

這便是我們的貢獻。

Now I just want to wrap up,

現在，我只是想結尾，

and as I was mentioning earlier

如我以前提到,

of course I have a lot of other data

當然我還有很多其他的數據，

if you're interested, but I just wanted to give

如果你有興趣知道，但我只是想

this sort of basic idea

給你看看這種

of being able to communicate

能夠與大腦思路溝通

with the brain in its language, and

的基本語言，並能夠做到

the potential power of being able to do that.

這一點的潛在力量。

So it's different from the motor prosthetics

因此，這是不同於用義馬達的設置

where you're communicating from the brain

從大腦到設備的通信假肢。

to a device. Here we have to communicate

我們這個的溝通

from the outside world

是來自外部世界，

into the brain and be understood,

進入腦和被理解，

and be understood by the brain.

被大腦理解。

And then the last thing I wanted

最後我想說的一件事，

to say, really, is to emphasize

真的，是要強調

that the idea generalizes.

推廣的想法。

So the same strategy that we used

因而使我們找到代碼的方法，

to find the code for the retina we can also

我們也可以使用視網膜同樣的策略

use to find the code for other areas,

找到在其他地區的代碼，

for example, the auditory system and

例如，聽覺系統

the motor system, so for treating deafness

和機動系統，來治療耳聾

and for motor disorders.

和運動障礙。

So just the same way that we were able to

因此，以相同的方式，我們能夠

jump over the damaged

飛越過在視網膜

circuitry in the retina to get to the retina's

損壞了的輸出細胞的線路，

output cells, we can jump over the

我們可以飛越過

damaged circuitry in the cochlea

在耳蝸裡被損壞的

to get the auditory nerve,

聽覺神經的線路，

or jump over damaged areas in the cortex,

或飛越過大腦皮層下破壞的，

in the motor cortex, to bridge the gap

運動皮層的地區，架橋渡過

produced by a stroke.

中風所產生的影響。

I just want to end with a simple

我只是想以一個簡單的信息結束，

message that understanding the code

理解代碼是真的，

is really, really important, and if we

真的很重要，如果我們能

can understand the code,

理解代碼，

the language of the brain, things become

大腦的語言，以往似乎

possible that didn't seem obviously

明顯不可能的事情

possible before. Thank you.

便會成為有可能。謝謝。

(Applause)

(掌聲)