A test for Parkinson's with a phone call
-
0:01 - 0:03So, well, I do applied math,
-
0:03 - 0:04and this is a peculiar problem
-
0:04 - 0:06for anyone who does applied math, is that
-
0:06 - 0:08we are like management consultants.
-
0:08 - 0:10No one knows what the hell we do.
-
0:10 - 0:13So I am going to give you some -- attempt today
-
0:13 - 0:15to try and explain to you what I do.
-
0:15 - 0:18So, dancing is one of the most human of activities.
-
0:18 - 0:22We delight at ballet virtuosos and tap dancers
-
0:22 - 0:23you will see later on.
-
0:23 - 0:26Now, ballet requires an extraordinary level of expertise
-
0:26 - 0:29and a high level of skill,
-
0:29 - 0:31and probably a level of initial suitability
-
0:31 - 0:33that may well have a genetic component to it.
-
0:33 - 0:36Now, sadly, neurological disorders such as Parkinson's disease
-
0:36 - 0:39gradually destroy this extraordinary ability,
-
0:39 - 0:41as it is doing to my friend Jan Stripling, who was
-
0:41 - 0:44a virtuoso ballet dancer in his time.
-
0:44 - 0:47So great progress and treatment has been made over the years.
-
0:47 - 0:50However, there are 6.3 million people worldwide
-
0:50 - 0:53who have the disease, and they have to live with
-
0:53 - 0:56incurable weakness, tremor, rigidity
-
0:56 - 0:58and the other symptoms that go along with the disease,
-
0:58 - 1:00so what we need are objective tools
-
1:00 - 1:03to detect the disease before it's too late.
-
1:03 - 1:06We need to be able to measure progression objectively,
-
1:06 - 1:09and ultimately, the only way we're going to know
-
1:09 - 1:11when we actually have a cure is when we have
-
1:11 - 1:14an objective measure that can answer that for sure.
-
1:14 - 1:17But frustratingly, with Parkinson's disease
-
1:17 - 1:20and other movement disorders, there are no biomarkers,
-
1:20 - 1:22so there's no simple blood test that you can do,
-
1:22 - 1:24and the best that we have is like
-
1:24 - 1:26this 20-minute neurologist test.
-
1:26 - 1:28You have to go to the clinic to do it. It's very, very costly,
-
1:28 - 1:31and that means that, outside the clinical trials,
-
1:31 - 1:34it's just never done. It's never done.
-
1:34 - 1:37But what if patients could do this test at home?
-
1:37 - 1:39Now, that would actually save on a difficult trip to the clinic,
-
1:39 - 1:43and what if patients could do that test themselves, right?
-
1:43 - 1:45No expensive staff time required.
-
1:45 - 1:47Takes about $300, by the way,
-
1:47 - 1:49in the neurologist's clinic to do it.
-
1:49 - 1:51So what I want to propose to you as an unconventional way
-
1:51 - 1:53in which we can try to achieve this,
-
1:53 - 1:55because, you see, in one sense, at least,
-
1:55 - 1:58we are all virtuosos like my friend Jan Stripling.
-
1:58 - 2:02So here we have a video of the vibrating vocal folds.
-
2:02 - 2:05Now, this is healthy and this is somebody making speech sounds,
-
2:05 - 2:08and we can think of ourselves as vocal ballet dancers,
-
2:08 - 2:11because we have to coordinate all of these vocal organs
-
2:11 - 2:13when we make sounds, and we all actually
-
2:13 - 2:15have the genes for it. FoxP2, for example.
-
2:15 - 2:18And like ballet, it takes an extraordinary level of training.
-
2:18 - 2:20I mean, just think how long it takes a child to learn to speak.
-
2:20 - 2:23From the sound, we can actually track
-
2:23 - 2:25the vocal fold position as it vibrates,
-
2:25 - 2:28and just as the limbs are affected in Parkinson's,
-
2:28 - 2:30so too are the vocal organs.
-
2:30 - 2:32So on the bottom trace, you can see an example of
-
2:32 - 2:34irregular vocal fold tremor.
-
2:34 - 2:35We see all the same symptoms.
-
2:35 - 2:38We see vocal tremor, weakness and rigidity.
-
2:38 - 2:40The speech actually becomes quieter and more breathy
-
2:40 - 2:42after a while, and that's one of the example symptoms of it.
-
2:42 - 2:45So these vocal effects can actually be quite subtle,
-
2:45 - 2:48in some cases, but with any digital microphone,
-
2:48 - 2:51and using precision voice analysis software
-
2:51 - 2:53in combination with the latest in machine learning,
-
2:53 - 2:55which is very advanced by now,
-
2:55 - 2:58we can now quantify exactly where somebody lies
-
2:58 - 3:01on a continuum between health and disease
-
3:01 - 3:03using voice signals alone.
-
3:03 - 3:06So these voice-based tests, how do they stack up against
-
3:06 - 3:08expert clinical tests? We'll, they're both non-invasive.
-
3:08 - 3:12The neurologist's test is non-invasive. They both use existing infrastructure.
-
3:12 - 3:15You don't have to design a whole new set of hospitals to do it.
-
3:15 - 3:17And they're both accurate. Okay, but in addition,
-
3:17 - 3:20voice-based tests are non-expert.
-
3:20 - 3:22That means they can be self-administered.
-
3:22 - 3:25They're high-speed, take about 30 seconds at most.
-
3:25 - 3:27They're ultra-low cost, and we all know what happens.
-
3:27 - 3:30When something becomes ultra-low cost,
-
3:30 - 3:32it becomes massively scalable.
-
3:32 - 3:36So here are some amazing goals that I think we can deal with now.
-
3:36 - 3:38We can reduce logistical difficulties with patients.
-
3:38 - 3:40No need to go to the clinic for a routine checkup.
-
3:40 - 3:43We can do high-frequency monitoring to get objective data.
-
3:43 - 3:47We can perform low-cost mass recruitment for clinical trials,
-
3:47 - 3:49and we can make population-scale screening
-
3:49 - 3:51feasible for the first time.
-
3:51 - 3:53We have the opportunity to start to search
-
3:53 - 3:56for the early biomarkers of the disease before it's too late.
-
3:56 - 3:59So, taking the first steps towards this today,
-
3:59 - 4:01we're launching the Parkinson's Voice Initiative.
-
4:01 - 4:03With Aculab and PatientsLikeMe, we're aiming
-
4:03 - 4:05to record a very large number of voices worldwide
-
4:05 - 4:09to collect enough data to start to tackle these four goals.
-
4:09 - 4:10We have local numbers accessible to three quarters
-
4:10 - 4:12of a billion people on the planet.
-
4:12 - 4:15Anyone healthy or with Parkinson's can call in, cheaply,
-
4:15 - 4:17and leave recordings, a few cents each,
-
4:17 - 4:19and I'm really happy to announce that we've already hit
-
4:19 - 4:23six percent of our target just in eight hours.
-
4:23 - 4:27Thank you. (Applause)
-
4:27 - 4:33(Applause)
-
4:33 - 4:36Tom Rielly: So Max, by taking all these samples of,
-
4:36 - 4:39let's say, 10,000 people,
-
4:39 - 4:42you'll be able to tell who's healthy and who's not?
-
4:42 - 4:44What are you going to get out of those samples?
-
4:44 - 4:46Max Little: Yeah. Yeah. So what will happen is that,
-
4:46 - 4:47during the call you have to indicate whether or not
-
4:47 - 4:49you have the disease or not, you see. TR: Right.
-
4:49 - 4:51ML: You see, some people may not do it. They may not get through it.
-
4:51 - 4:54But we'll get a very large sample of data that is collected
-
4:54 - 4:57from all different circumstances, and it's getting it
-
4:57 - 4:59in different circumstances that matter because then
-
4:59 - 5:02we are looking at ironing out the confounding factors,
-
5:02 - 5:05and looking for the actual markers of the disease.
-
5:05 - 5:07TR: So you're 86 percent accurate right now?
-
5:07 - 5:08ML: It's much better than that.
-
5:08 - 5:10Actually, my student Thanasis, I have to plug him,
-
5:10 - 5:12because he's done some fantastic work,
-
5:12 - 5:16and now he has proved that it works over the mobile telephone network as well,
-
5:16 - 5:19which enables this project, and we're getting 99 percent accuracy.
-
5:19 - 5:21TR: Ninety-nine. Well, that's an improvement.
-
5:21 - 5:23So what that means is that people will be able to —
-
5:23 - 5:25ML: (Laughs)
-
5:25 - 5:27TR: People will be able to call in from their mobile phones
-
5:27 - 5:30and do this test, and people with Parkinson's could call in,
-
5:30 - 5:33record their voice, and then their doctor can check up
-
5:33 - 5:35on their progress, see where they're doing in this course of the disease.
-
5:35 - 5:36ML: Absolutely.
-
5:36 - 5:38TR: Thanks so much. Max Little, everybody.
-
5:38 - 5:43ML: Thanks, Tom. (Applause)
- Title:
- A test for Parkinson's with a phone call
- Speaker:
- Max Little
- Description:
-
Parkinson’s disease affects 6.3 million people worldwide, causing weakness and tremors, but there's no objective way to detect it early on. Yet. Applied mathematician and TED Fellow Max Little is testing a simple, cheap tool that in trials is able to detect Parkinson's with 99 percent accuracy -- in a 30-second phone call.
- Video Language:
- English
- Team:
- closed TED
- Project:
- TEDTalks
- Duration:
- 06:04
Thu-Huong Ha edited English subtitles for A test for Parkinson's with a phone call | ||
Thu-Huong Ha approved English subtitles for A test for Parkinson's with a phone call | ||
Thu-Huong Ha edited English subtitles for A test for Parkinson's with a phone call | ||
Morton Bast accepted English subtitles for A test for Parkinson's with a phone call | ||
Morton Bast edited English subtitles for A test for Parkinson's with a phone call | ||
Joseph Geni added a translation |