ADM-201 dump PMP dumps pdf SSCP exam materials CBAP exam sample questions

一种检测癫痫发作的人工智能手表 – 译学馆
未登陆,请登陆后再发表信息
最新评论 (0)
播放视频

一种检测癫痫发作的人工智能手表

An AI smartwatch that detects seizures | Rosalind Picard

这是亨利 一个可爱的男孩
This is Henry, a cute boy.
当他三岁时
And when Henry was three,
他妈妈发现他有些热性癫痫的症状
his mom found him having some febrile seizures.
热性癫痫是发作时伴随着发烧的癫痫病
Febrile seizures are seizures that occur when you also have a fever.
医生说:“不用太担心 小孩长大点这种病就好了”
And the doctor said: “don’t worry too much,kids usually outgrow these.”
当他四岁时 他患上了痉挛性癫痫
When he was four, he had a convulsive seizure,
会失去意识并抖动的那种
the kind that you lose consciousness and shake,
全身强直阵挛发作
a generalized tonic-clonic seizure.
当癫痫诊断书还在邮寄途中时
And while the diagnosis of epilepsy was in the mail,
某天早上 亨利的妈妈去叫他起床
Henry’s mom went to get him out of bed one morning,
当她走进他的房间
and as she went in his room,
她发现他身体冰凉 没有生命迹象
she found his cold, lifeless body.
亨利死于SUDEP
Henry died of SUDEP,
即癫痫猝死症
sudden unexpected death in epilepsy.
我好奇在座有多少人听过癫痫猝死症
I’m curious how many of you have heard of SUDEP.
各位都是受过良好教育的观众
This is a very well-educated audience,
但我只看到几个人举手
and I see only a few hands.
癫痫猝死症指的是健康的癫痫患者死亡时
SUDEP is when an otherwise healthy person with epilepsy dies
无法在尸检中找到病因
and they can’t attribute it to anything they can find in an autopsy.
每七到九分钟就有一个人发生癫痫猝死症
There is a SUDEP every seven to nine minutes.
平均每场TED演讲就有两人
That’s on average two per TED Talk.
正常的大脑会产生脑电波
Now, a normal brain has electrical activity.
你可以从这张大脑图片上看到脑电波
You can see some of the electrical waves coming out of this picture of a brain here.
这是脑电图描记器可以直接读到的
And these should look like typical electrical activity
典型的脑电活动
that an EEG could read on the surface.
当癫痫发作时
When you have a seizure,
就会有异常的脑电活动
it’s a bit of unusual electrical activity.
这可能就是病灶所在
And it can be focal,
它可以只发生在大脑的一小块区域
It can take place in just a small part of your brain.
当它发生时
When that happens,
你可能会有奇怪的感觉
you might have a strange sensation.
在座的观众当中可能有一些人现在就有这个状况
One could be happening several could be happening here in the audience right now,
而你旁边的人可能完全不知道
and the person next to you might not even know.
然而 如果你脑中癫痫发作的病灶从小小的火苗
However, if you have a seizure where that little brush fire
扩散到像森林大火那样
spreads like a forest fire over the brain,
癫痫就会变成全身性的
then it generalizes.
全身性的发作
And that generalized seizure
会令你失去意识
takes your consciousness away
并让你抽搐
and cause you to convulse.
在美国 每年因癫痫猝死的人数
There are more SUDEPs in the United States every year
高于因婴儿猝死综合症(死亡的人数)
than sudden infant death syndrome.
在座有多少人听过婴儿猝死综合症?
Now, how many of you have heard of sudden infant death syndrome?
好的 几乎人人都举手了
All right,pretty much every hand goes up.
那么 这是怎么回事?
So, what’s going on here?
为什么这种疾病这么普遍
Why is this so much more common
然而大家却没听说过它呢?
and yet people haven’t heard of it?
你要如何预防它呢?
And what can you do to prevent it?
科学指出有两个方法
All the two things scientifically shown,
可以预防或降低癫痫猝死症的风险
that prevent or reduce the risk of SUDEP.
第一 谨遵医嘱 按时吃药
The first is, follow your doctor’s instructions, take your medications.”
三分之二的癫痫症患者
Two-thirds of people who have epilepsy
都能用药物来控制
get it under control with their medications.
能降低癫痫猝死症风险的第二个方法就是 陪伴
The second thing that reduces the risk of SUDEP is companionship.
也就是当你发作的时候 要有人在身边
It’s having somebody there at the time that you have a seizure.
虽然大部分人没听过癫痫猝死症
Now, SUDEP, even through most of you have never heard of it,
但它其实是所有神经疾病中最有可能致死的第二名
is actually the number two cause of years of potential life lost of all neurological disorders.
纵轴是死亡人数
The vertical axis is the number of deaths
乘上剩下的生命时长
times the remaining life span,
那么 越高表示影响越糟糕
so, higher is much worse impact.
然而 癫痫猝死症不像其他疾病
SUDEP, however, unlike these others,
人们能做点什么
is something that people right here
来压制住它
could do something to push that down.
为什么罗兹•皮卡德
Now, what is a Roz Picard,
人工智能研究者 要在这里跟大家谈癫痫猝死症?我不是神经病学家
AI researcher, doing here telling you about SUDEP, right? I’m not a neurologist.
我在麻省理工媒体实验室
When I was working at the Media Lab
研究情绪测量时
on measurement of emotion,
试图让我们的机器更好的判断识别我们的情绪
trying to make our machinesmore intelligent about our emotions,
我们开始努力去测量压力
we started doing a lot of work measuring stress.
我设置了许多感应器
We build lots of sensors
用许多不同的方式来测量压力
that measured it in lots of different ways.
但其中有种方式很特别
But one of them in particular,
是用电信号这种古老的方式
grew out of some of this very old work
来测量掌心流汗
with measuring sweaty palms with an electrical signal.
这是皮肤电导的信号
This is a signal of skin conductance
大家都知道 当你紧张时 这种信号会上升
that’s known to go up when you get nervous,
但结果发现
but it turns out also goes up with
还有许多其他有趣的情况也会让它上升
a lot of other interesting conditions.
但手上连接电线来测量实在不太方便
But measuring it with wires on our hand is really inconvenient.
所以 我们在麻省理工媒体实验室里研究出了许多其他方法来做这件事
So we invented a bunch of other ways of doing this at the MIT Media Lab.
有了这些可穿戴的装备
And with these wearables,
我们开始能取得一周七天 一天二十四小时的高质量临床资料 这是前所未有的
we started to collect the first-ever clinical quality data 24-7.
这张图上的是
Here’s a picture of what that looked like
第一次从麻省理工学院的学生那里收集的
the first time an MIT student collected
一周七天 一天二十四小时的手腕上的皮肤电导资料
skin conductance on the wrist 24-7.
咱们把它放大一点
Let’s zoom in a little bit here.
从左到右是二十四小时的情况
What you see is 24 hours from left to right,
图上的是两天的资料
and here is two days of data.
首先 让我们惊讶的是
And first, what surprised us was sleep
睡觉时是一天中峰值发生的时间
was the biggest peak of the day.
听起来是出故障了 对吧?
Now, that’s sounds broken, right?
怎么会呢 我是说 你睡着时是很平静的
How? I mean you’re calm when you’re asleep,
那这是怎么回事?
so what’s going on here?
结果发现 睡眠期间的生理机能
Well, it turns out that our physiology during sleep
和清醒期间非常不同
is very diffrent than our physiology during awake,
虽然对于为什么一天之中的峰值
and while there’s still a bit of a mystery
通常出现在睡眠期间 仍然有点难以理解
why these peaks are usually the biggest of the day during sleep.
但我们现在相信
we now believe they’re related to
它们和记忆的巩固
memory consolidation
及睡眠期间记忆的形成有关
and memory formation during sleep.
还有一些发现 完全在我们的预料之中
We also saw things that were exactly what we expected.
当麻省理工学院的学生在实验室努力或在做功课时
When an MIT student is working hard in the lab or on homeworks,
不仅有情绪压力
there is not only emotional stress,
还有认知负荷
but there’s cognitive load,
结果表明 认知负荷
and it turns out that cognitive load,
认知努力 心理参与
cognitive effort, mental engagement,
对学习某件事物感到兴奋
excitement about learning something
上述这些也都会造成信号上升
those things also make the signal go up.
很遗憾 让我们麻省理工学院教授们尴尬的是
Unfortunately, the embarrassment of we MIT professors,
一天的低点出现在课堂上
the low point every day is classroom activity.
现在 各位在这里看到的只是一个人的资料
Now, I am just showing you one person’s data here,
但遗憾的是 这也是个普遍现象
but this, unfortunately is true in general.
这条防汗带
This sweatband
里面有自制的皮肤电导感测器
has inside it a homebuilt skin-conductance sensor.
有一天 我们的一位大学生来敲我的门
And one day, one of our undergrads knocked on my door
那时是十二月份学期末
right at the end of the December semester,
他说:“皮卡德教授 能借给我一个腕带传感器吗?
and he said:”Professor Picard, can I please borrow one of your wristband sensors?
我弟弟有自闭症
My little brother has autism,
他不能说话 我想看看是什么使他紧张”
he can’t talk, and I want to see what’s streesing him out.”
我说:“当然可以 别只拿一条 拿两条 因为它们很容易坏”
And I said,” Sure, in fact, don’t just take one, take two because they broke easily back then.”
于是 他把腕带传感器带回了家
So he took them home,
并给他弟弟戴上
he put them on his little brother.
我回到麻省理工大学并在笔记本上看着数据
Now, I was back in MIT, looking at the data on my laptop.
第一天
And the first day,
我想着 “哦 很奇怪 他把传感器同时放在了两个胳膊腕上
I thought, “um, that’s odd, he put them on both wrists
而不是等一条坏了再换上另一条
instead of waiting for one to break.
好吧 别听我的指示”
Ok, fine, don’t follow my instructions.”
我很开心他没听我的
I’m glad he didn’t.
第二天——扫兴 看起来和课堂上一样
Second day–chill. Looked like classroom activity.
又过了几天
A few more days ahead.
接下来的一天
The next day,
一条腕上的信号平稳
one wrist signal was flat
另一条腕上出现了我见过的峰顶
and the other had the biggest peak I’ve ever seen,
我在想“发生了什么?
and I thought, “what’s going on?
我们在MIT采用了所有可以想到的办法给人们施压
We’ve stressed people out at MIT every way imaginable.
也从未见过这么高的峰顶”
I’ve never seen a peak this big.”
而且只有一边有这种变化
And it was only on one side.
怎么会只有一边有压力而另一边没事呢?
How can you be stressed on one side of your body and not the other?
所以我认为 其中一个或这两个感应器都出了问题
So I thought one or both sensors must be broken.
现在经过训练 我成了电镀工程师
Now, I’m an electro engineer by training,
于是 我开始采用一切措施去调试感应器
so I started a whole bunch of stuff to try to debug this,
长话短说
and long story short,
我没能使之重现
I could not reproduce this.
所以我采取老办法进行调试
So I resorted to old-fashioned debugging.
假期的时候 我在家给学生打了电话
I called the student at home on vacation.
“嗨 你弟弟怎么样了?
“Hi, how’s your little brother?
圣诞节过得怎么样?
How’s your Christmas?
嘿 你对他身上发生的情况有什么看法?”
Hey, do you have any idea what happened to him?”
我在数据中给出了这个特定的日期和时间
And I give this particular date and time in data.
他说
And he said,
“我不知道 我要查一下日记”
“I don’t know, I’ll check the diary.”
日记?一个MIT学生还坚持写日记?
Diary? An MIT student keeps a diary?
所以我等着 他把准确的数据和时间记录拿来了
So I waited and he came back had exact data and time,
他说“没错 在他发病之前 一切都很正常”
and he says, “that was right before he had a grand mal seizure.”
那个时候 我对癫痫并不了解
Now, at the time, I didn’t know anything about epilepsy,
于是做了大量的研究
and did a bunch of research,
想到另一个学生的父亲是波士顿儿童医院的神经外科的主治医生
realized that another student’s dad is chief of neurosurgery at Children’s Hospital Boston,
我鼓足勇气打了电话给马德森医生
screwed up my courage and called Dr. Joe Madsen.
“嗨 马德森医生 我叫皮卡德
“Hi, Dr. Madsen, my name’s Rosalind Picard.
有没有可能某些人
Is it possible somebody could have
在突然发病前
a huge sympathetic nervous system surge
交感神经系统出现20分钟的巨大波动
that’s what drives the skin conductance
促使皮肤电导系数突然高涨”
20 minutes before a seizure?”
他说:“可能不会”
And he says:”probably not.”
他说“这很有趣 我们有个人在发病前
He says, “It’s interesting,we’ve had people whose hair stands on end on one arm
一只胳膊上的汗毛突然竖了20分钟
20 minutes before a seizure.”
我在想“一只胳膊?”
And I’m like, “On one arm?”
起初 我不想告诉他
I did’t want to tell him that, initially,
因为我认为这太不可思议了
because I thought this was too ridiculous.
他向我解释了大脑是如何产生这种反应的
He explained how this could happen in the brain,
他对此很感兴趣
and he got interested.
我把数据向他展示了 我们制作了大量的装置
I showed him the data we made a whole bunch more devices,
确保安全
got them safety certified.
90个家庭参加了这项研究
90 families were being enrolled in a study,
包括他们的孩子 七天每天24小时
all with children who were going to be monitored 24-7
用最标准的EEG放在他们的头上
with gold-standard EEG on their scalp
对他们的大脑活动进行监控
for reading the brain activity,
通过视频观看大脑活动
video to watch the behavior,
心电图还有EDA
electrocardiogram ECG and now EDA,
皮肤电活动
electrodermal activity,
来观察周围是否有一些
to see if there was something in this periphery
我们可以很容易看出跟癫痫发作相关的东西
that we could easily pick up, related to a seizure.
我们发现 在第一批癫痫大发作中
We found, in 100% of the first batch of grand mal seizures,
所有的人都出现了这种巨大的皮肤电导反应
this whopper of responses in the skin conductance.
中间蓝色的部分 这个男孩睡觉时
The blue in the middle, the boy’s sleep,
通常处于一天中的最高峰
is usually the biggest peak of the day.
你看这儿三次癫痫发作
These three seizures you see here
就像森林中长出的红色的树
are popping out of the forest like reswood trees.
进一步来讲 当你将顶部皮肤电导和
Futhermore, when you couple the skin conductance at the top
腕部的运动连在一起的时候
with the movement from the wrist,
你会得到很多数据
and you get lots of data
如果把训练机器和AI放在上面
and train machine learning and AI on it,
就能建立一个全自动的监测这些形式的人工智能
you can build an automated AI that detects these patterns,
这样比震动探测器监测的更好
much better than just a shake detector can do.
我们意识到要解决这个问题
So we realized that we needed to get this out,
通过Ming-ZherPoh的博士研究
and with PHD work of Ming-ZherPoh
以及后来Empatica的巨大改进
and later great improvements by Empatica,
这使我们监测癫痫的数据精确多了
this has made progress and the seizure drtection is much more accurate.
在这期间 我们也学到了一些关于SUDEP的其他的东西
But we also learned some other things about SUDEP during this.
一个是 我们知道SUDEP
One thing we learned is that SUDEP is
在一般强直性阵挛发作后很少发生
while it’s rare after a generalized tonic-clonic seizure,
但这种类型发作后 最有可能发生
that’s when it’s most likely to happen after that type.
当其发作时 不会出现这种情况
And when it happens, it doesn’t happen during the seizure,
突然发病后通常不会立刻发生SUDEP
and it doesn’t usually happen immediately afterwards,
但一会儿之后
but imediately afterwards,
当这个人静止并安静下来之后
when the person just seeems very still and quiet,
他们可们进入另一个阶段
they may go into another phase,
呼吸停止
where the breathing stops,
呼吸停止之后
and then after the breathing stops,
然后 心脏停止跳动
later the heart stops.
所以人们需要在那儿待一段时间
So there’s some time to get somebody there.
我们也了解到大脑深处有一个区域
We also learned that there is a region
叫做杏仁孔
deep in the brain called the amygdale,
我们情绪实验一直在研究这个区域
which we had been studying in our emotion research a lot.
我们有两个杏仁孔
We have two amygdalas.
如果你刺激右边的杏仁孔
And if you stimulate the right one,
你会看到右边的皮肤电导有很大的反应
you get a big right skin conductance response.
你必须立刻签字同意做开颅术进行治疗
Now, you have to sign up right now for a craniotomy to get this done,
我们不确定要求志愿者具体做什么
not exactly something we’re going to volunteer to do,
但这导致右面皮肤电导产生强烈的反应
but it cause a big right skin conductance response.
刺激左边 手掌会有很强烈的皮肤电导反应
Stimulate the left one, big left skin conductance response on the palm.
进一步讲 当某人刺激你的杏仁孔
And furthemore, when somebody stimulates your amygdala
这时你就坐在那儿 可能一直在工作
while you’re sitting there and you might just be working,
你不会表现出任何痛苦的迹象
you don’t show any signs of distress,
但你停止了呼吸
but you stop breathing.
你不会再动了
And you don’t start again
直到某人刺激你
until somebody stimulates you.
“嘿 罗兹 你在吗?”
“Hey, Roz, are you there?”
你张开嘴说话
And you open your mouth to talk.
当你想要呼吸说话的时候
As you take that breath to speak,
你开始再次呼吸
you start breathing again.
所以我们顶着压力开始工作
So we had started with work on stress,
这样使我们建立许多传感器
which had enabled us to build lots of sensors
置于一处 给我们传送高质量的足够的数据
that were gathering high quality enough data
我们可以离开实验室 开始在野外收集数据
that we could leave the lab and start to get this in the wild;
我们意外地发现 突然发病是个谎言
accidentally found a whopper of a response with the seizure,
神经病学的活动能比传统的压力因素
neurological activation that can cause a much bigger response
造成更大的反应
than traditional stressors;
许多医院的合作企业
lots of partnership with hospitals
和癫痫控制科室
and an epilepsy monitoring unit,
尤其是波士顿儿童医院和布里格姆
especially Children’s Hospital Boston and the Brigham;
机器和还有上面这个人工智能
and machine learning and AI on top of this
能抓取并收集很多数据
to take and collect lots more data
这些东西帮助我们理解这些事
in service of trying to understand these events
如果我们能预防SUDEP(就更好了)
and if we could prevent SUDEP.
这就是现在被Empatica商业化的
This is now commercialized by Empatica,
一家新兴的 我有幸与人合作的企业
a start-up that I had the privilege to cofound,
这个团队完成了一项惊人的工作
and the team there has done an amazing job improving the technology
改善了技术 做出了一种非常漂亮的传感器
to make a very beautiful sensor
不仅可以报时
that not only tells the time and does
计步 睡眠 做其他的事情
steps and sleep and all that good stuff,
而且人工智能和机器可以实时
but this is running real-time AI and machine learning
监测全身强直性阵挛性癫痫
to detect generalized tonic-clonic seizures
如果我癫痫发作失去意识
and send an alert for help
它会发射信号进行求助
if I were t have a seizure and lose consciousness.
这个东西得到了FDA的认证
This just got FDA-approved
它是第一个在神经学上
as the first smartwatch
得到认证的智能手表
to get approved in neurology.
现在 下一个情景是使我的皮肤电导上升的东西
Now, the next slide is what made my skin conductance go up.
一天早晨 我在检查邮件时
One mornig, I’m checking my email
看见一位妈妈给我发来的一个故事
and I see a story from a mom
她说她在淋浴时
who said she was in the shower,
她的手机在淋浴旁边的柜台上
and her phone was on the counter by the shower,
她说她的女儿可能需要她的帮助
and said her daughter might need her help.
于是她停止洗澡跑向她女儿的卧室
So she interrrupts her shower and goes running to her daughter’s bedroom,
她发现她的女儿脸向着床
and she finds her daughter facedown in bed,
脸色发青停止了呼吸
blue and not breathing.
她将女儿翻过来
She flips her over human stimulation
她的女儿恢复了呼吸
and her daughter takes a breath,
脸色恢复了红润并逐渐好转
ane another breath,and her daughter turns pink and is fine.
我想我读完这封邮件时 脸色变得苍白
I think I turned white reading this email.
我的第一反应是“不 这不对
My first reaponse is, “Oh no, it’s not perfect.
蓝牙会坏 电池会没电
The Blurtooth could break, the battery could die.
一切都会出问题 不要依赖这个”
All the things could go wrong. Don’t rely on this.”
她说“没关系 我知道没有完美的技术
And she said, “It’s OK. I know no technology is perfect.
我们没有人能一直在那儿看着
None of us can always be there all the time.
但这个 这个装置加上AI
But this, this device plus AI
使我及时赶到救了我女儿”
enabled me to get there in time to save my daughter’s life.”
现在 我一直在强调孩子们
Now, I’ve been mentioning children,
但是SUDEP高发人群是
but SUDEP peaks, actually, among people
20岁30岁40岁的人群
in their 20s, 30s and 40s,
我要提出的下一条
and the next line I’m going to put up
可能使一些人不舒服
is probably going to make some people uncomfortable,
但大家要知道 如果这条建议扩展到你认识的某个人
but it’s less uncomfortale than we’ll all be
我们都会舒服很多
if this list is extended to somebody you know.
这会发生在你认识的某个人身上吗?
Could this happen to somebody you know?
我提一个令人不舒服的问题的原因是
And the reason I bring up this uncomfortable question
因为每26个人中就有1个
is because one in 26 if you
在某种程度上患有癫痫
will have epilepsy at some point,
从我一直的研究来看
and from what I’ve been learning,
患有癫痫的人经常不会告诉他们的朋友和邻居
people with epilepsy often don’t tell their friends and their neighbors that they have it.
所以如果你愿意
So it you’re willing
让他们使用AI或者
to let them use an AI or whatever
无论什么时候当你需要它 它都会满足你的需要
to summon you in a moment of possible need,
如果你可以让他们知道
if you would let them know that,
你会在他们的生活中起到重大的影响
you could male a diffrence in their life.
为什么要努力建立AI呢?
Why do all this hard work to build AIs?
这有几个原因:一个是娜塔莎
Couple of reasons here: one is Natashe,
这个活下来的女孩
the girl who lived,
她的家人想让我告诉大家她的名字
and her family wanted me to tell you her name.
另一个是她的家人
Another is her family
和外面那些很好的人
and the wonderful people out there
他们是想支持
who want to be there to support people
那些过去一向别人提及自己的状况
who have conditions that they’ve felt uncomfortable
就感到不舒服的人
in the past mentioning to others.
另一个原因就是在座的各位
And the other reason is all of you,
因为我们有机会
because we have the opportunity
在未来改进AI技术
to shape the future AI.
我们其实可以改变它
We can actually change it,
因为是我们发明的它
because we are rhe ones building it.
所以我们来利用AI
So let’s build AI
让每个人的生活变得更美好吧
that makes everybody’s lives better.
谢谢大家
Thank you.

发表评论

译制信息
视频概述

本文主要讲了检测癫痫发作的一款人工智能手表

听录译者

甪里

翻译译者

孙卫卫

审核员

审核员 V

视频来源

https://www.youtube.com/watch?v=YkYdxjaqt58

相关推荐