Can AI Catch What Doctors Miss? | Eric Topol | TED

168,879 views ・ 2023-12-09

TED


μ•„λž˜ μ˜λ¬Έμžλ§‰μ„ λ”λΈ”ν΄λ¦­ν•˜μ‹œλ©΄ μ˜μƒμ΄ μž¬μƒλ©λ‹ˆλ‹€.

λ²ˆμ—­: Seongjae Hwang κ²€ν† : DK Kim
00:05
I've had the real fortune of working at Scripps Research
0
5961
3336
μ €λŠ” 정말 운이 μ’‹κ²Œ μ§€λ‚œ 17λ…„ λ™μ•ˆ 슀크립슀 λ¦¬μ„œμΉ˜μ—μ„œ μΌν–ˆμŠ΅λ‹ˆλ‹€.
00:09
for the last 17 years.
1
9339
1835
00:11
It's the largest nonprofit biomedical institution in the country.
2
11174
5255
λ―Έκ΅­μ—μ„œ κ°€μž₯ 큰 λΉ„μ˜λ¦¬ μƒμ˜ν•™ μ—°κ΅¬μ†Œμ£ .
00:16
And I've watched some of my colleagues,
3
16972
2419
μ €λŠ” 제 λ™λ£Œλ“€ λͺ‡ λͺ…이 삼사 년이 κ±Έλ €μ„œ
00:19
who have spent two to three years
4
19391
2294
00:21
to define the crystal 3-D structure of a protein.
5
21685
3253
3차원 λ‹¨λ°±μ§ˆ κ²°μ • ꡬ쑰λ₯Ό ν™•μΈν•˜λŠ” 것을 μ§€μΌœλ΄€μŠ΅λ‹ˆλ‹€.
00:26
Well, now that can be done in two or three minutes.
6
26231
3420
μ΄μ œλŠ” 이삼 λΆ„ μ•ˆμ— 이 μž‘μ—…μ„ μ™„λ£Œν•  수 μžˆμŠ΅λ‹ˆλ‹€.
00:29
And that's because of the work of AlphaFold,
7
29693
2419
λ”₯λ§ˆμΈλ“œμ—μ„œ νŒŒμƒλœ μ•ŒνŒŒν΄λ“œ 덕뢄인데
00:32
which is a derivative of DeepMind, Demis Hassabis and John Jumper,
8
32112
6882
λ”₯λ§ˆμΈλ“œμ˜ 데미슀 ν•˜μ‚¬λΉ„μŠ€, μ‘΄ μ νΌλŠ” 이 연ꡬ μ„±κ³Όλ‘œ
00:38
recognized by the American Nobel Prize in September.
9
38994
3503
μ§€λ‚œ 9μ›” 미ꡭ의 노벨상을 λ°›μ•˜μŠ΅λ‹ˆλ‹€.
00:42
What's interesting, this work,
10
42998
1918
이 μ—°κ΅¬μ—μ„œ ν₯미둜운 점은
00:44
which is taking the amino acid sequence in one dimension
11
44958
4588
1차원 μ•„λ―Έλ…Έμ‚° μ„œμ—΄μ„ κ°€μ§€κ³ 
00:49
and predicting the three-dimensional protein at atomic level,
12
49588
5296
μ›μž μˆ˜μ€€μ—μ„œ 3차원 λ‹¨λ°±μ§ˆμ„ μ˜ˆμΈ‘ν•˜λŠ” 이 연ꡬ가
00:54
[has] now inspired many other of these protein structure prediction models,
13
54926
5881
μˆ˜λ§Žμ€ λ‹€λ₯Έ λ‹¨λ°±μ§ˆ ꡬ쑰 예츑 λͺ¨λΈλ“€λ‘œ ν•˜μ—¬κΈˆ
01:00
as well as RNA and antibodies,
14
60849
2711
RNA 및 항체 예츑뿐 μ•„λ‹ˆλΌ
01:03
and even being able to pick up all the missense mutations in the genome,
15
63602
4838
심지어 μœ μ „μ²΄μ˜ λͺ¨λ“  λ―Έμ„Έ 변이λ₯Ό μ‹λ³„ν•˜κ³ 
01:08
and even being able to come up wit proteins
16
68481
4046
이전에 발λͺ…λœ 적이 μ—†κ³  μžμ—°μ— μ‘΄μž¬ν•˜μ§€ μ•ŠλŠ”
01:12
that have never been invented before, that don't exist in nature.
17
72569
3920
μƒˆλ‘œμš΄ λ‹¨λ°±μ§ˆκΉŒμ§€ 생각해 λ‚Ό 수 있게 ν•œ κ²ƒμž…λ‹ˆλ‹€.
01:16
Now, the only thing I think about this is it was a transformer model,
18
76990
3295
μ €λŠ” 이게 트랜슀포머 λͺ¨λΈμ΄λΌλŠ” μ μ—λ§Œ 관심이 κ°€λŠ”λ°
01:20
we'll talk about that in a moment,
19
80285
2169
그건 κ³§ μ΄μ•ΌκΈ°ν•˜κ² μŠ΅λ‹ˆλ‹€.
01:22
in this award, since Demis and John
20
82454
5046
이 상을 λ°›λŠ” κ³Όμ •μ—μ„œ 데미슀, μ‘΄κ³Ό κ³Όν•™μž 30λͺ… νŒ€μ€
01:27
and their team of 30 scientists
21
87542
1877
01:29
don't understand how the transformer model works,
22
89419
4004
트랜슀포머 λͺ¨λΈμ΄ μ–΄λ–»κ²Œ μž‘λ™ν•˜λŠ”μ§€ λͺ¨λ₯΄λ‹ˆκΉŒ
01:33
shouldn't the AI get an asterisk as part of that award?
23
93465
4963
μˆ˜μƒμž μ€‘μ—μ„œ AI에 κ°•μ‘° ν‘œμ‹œκ°€ λ˜μ–΄μ•Ό ν•˜μ§€ μ•Šμ„κΉŒμš”?
01:39
I'm going to switch from life science,
24
99262
2127
생λͺ… κ³Όν•™μ˜ 큰 κ³΅λ‘œμ— λŒ€ν•΄ μ§€κΈˆκΉŒμ§€ λ΄€λŠ”λ°
01:41
which has been the singular biggest contribution just reviewed,
25
101431
4129
이제 μ˜ν•™ μ΄μ•ΌκΈ°λ‘œ λ„˜μ–΄κ°€κ² μŠ΅λ‹ˆλ‹€.
01:45
to medicine.
26
105560
1335
01:47
And in the medical community,
27
107604
1501
μ˜ν•™κ³„μ—μ„œ λ³„λ‘œ μ΄μ•ΌκΈ°ν•˜μ§€ μ•ŠλŠ” μ£Όμ œλŠ”
01:49
the thing that we don't talk much about are diagnostic medical errors.
28
109105
6215
μ§„λ‹¨μ—μ„œ μƒκΈ°λŠ” μ˜ν•™μ  였λ₯˜μž…λ‹ˆλ‹€.
01:55
And according to the National Academy of Medicine,
29
115362
2669
κ΅­λ¦½μ˜ν•™μ›μ— λ”°λ₯΄λ©΄
우리 λͺ¨λ‘κ°€ 일생에 적어도 ν•œ λ²ˆμ€ κ²½ν—˜ν•˜κ²Œ 될 κ²ƒμž…λ‹ˆλ‹€.
01:58
all of us will experience at least one in our lifetime.
30
118031
3462
02:01
And we know from a recent Johns Hopkins study
31
121993
2294
쑴슀 ν™‰ν‚¨μŠ€ λŒ€ν•™μ˜ 졜근 연ꡬλ₯Ό 보면
02:04
that these errors have led to 800,000 Americans dead
32
124329
5672
이런 였λ₯˜ λ•Œλ¬Έμ— λ§€λ…„ 미ꡭ인 80만 λͺ…이
μ‚¬λ§ν•˜κ±°λ‚˜ μ‹¬κ°ν•œ μž₯μ• λ₯Ό μ–»λŠ”λ‹€κ³  ν•©λ‹ˆλ‹€.
02:10
or seriously disabled each year.
33
130043
3629
02:13
So this is a big problem.
34
133713
1502
λ”°λΌμ„œ μ΄λŠ” μ€‘λŒ€ν•œ λ¬Έμ œμž…λ‹ˆλ‹€.
02:15
And the question is, can AI help us?
35
135215
3211
κΆκΈˆν•œ 것은 κ³Όμ—° AIκ°€ 우리λ₯Ό λ„μšΈ 수 μžˆμ„κΉŒμž…λ‹ˆλ‹€.
02:18
And you keep hearing about the term β€œprecision medicine.”
36
138843
3295
μš°λ¦¬λŠ” 'μ •λ°€ 의료'λΌλŠ” 말을 계속 λ“£μŠ΅λ‹ˆλ‹€.
02:22
Well, if you keep making the same mistake over and over again, that's very precise.
37
142806
5589
뭐, 같은 μ‹€μˆ˜λ₯Ό 계속 λ°˜λ³΅ν•œλ‹€λ©΄ μ•„μ£Ό μ •λ°€ν•˜κΈ΄ ν•˜μ£ .
02:28
(Laughter)
38
148436
1168
(μ›ƒμŒ)
02:30
We don't need that,
39
150188
1168
그런 게 μ•„λ‹ˆλΌ μ •ν™•ν•˜λ©΄μ„œ μ •λ°€ν•œ μ˜ν•™μ΄ ν•„μš”ν•œ κ²λ‹ˆλ‹€.
02:31
we need accuracy and precision medicine.
40
151398
2669
02:34
So can we get there?
41
154442
1585
κ·Έλž˜μ„œ, κ°€λŠ₯ν• κΉŒμš”?
02:36
Well, this is a picture of the retina.
42
156486
2252
이건 망막 μ‚¬μ§„μž…λ‹ˆλ‹€.
이것이 첫 번째 μ€‘μš”ν•œ λ‹¨μ„œμ˜€μŠ΅λ‹ˆλ‹€.
02:39
And this was the first major hint,
43
159072
2961
02:42
training 100,000 images with supervised learning.
44
162075
4963
지도 ν•™μŠ΅μ„ 톡해 μ˜μƒ μ‹­λ§Œ 개λ₯Ό ν•™μŠ΅μ‹œμΌ°μ£ .
02:47
Could the machine see things that people couldn't see?
45
167080
4880
μ‚¬λžŒμ΄ λ³Ό 수 μ—†λŠ” 것듀을 기계가 λ³Ό 수 μžˆμ„κΉŒμš”?
02:52
And so the question was, to the retinal experts,
46
172919
2961
망막 μ „λ¬Έκ°€λ“€μ—κ²Œ 이런 μ§ˆλ¬Έμ„ λ˜μ‘ŒμŠ΅λ‹ˆλ‹€.
02:55
is this from a man or a woman?
47
175880
2002
이 망막 사진 주인은 λ‚¨μžμΌκΉŒμš”, μ—¬μžμΌκΉŒμš”?
02:58
And the chance of getting it accurate was 50 percent.
48
178717
3420
μ •ν™•ν•˜κ²Œ 맞힐 ν™•λ₯ μ€ 50%μ˜€μŠ΅λ‹ˆλ‹€.
03:02
(Laughter)
49
182137
1167
(μ›ƒμŒ)
03:03
But the AI got it right, 97 percent.
50
183346
3754
ν•˜μ§€λ§Œ AIλŠ” 97%λ₯Ό λ§žν˜”μŠ΅λ‹ˆλ‹€.
03:07
So that training,
51
187142
2043
심지어 이 ν•™μŠ΅μ—μ„œλŠ”
03:09
the features are not even fully defined of how that was possible.
52
189227
4171
ν•™μŠ΅ μžλ£Œκ°€ μ™„μ „νžˆ μ •μ˜λœ 것도 μ•„λ‹Œλ° 이게 μ–΄λ–»κ²Œ κ°€λŠ₯ν–ˆμ„κΉŒμš”?
03:14
Well that gets then to all of medical images.
53
194274
3086
λͺ¨λ“  의료 μ˜μƒμœΌλ‘œ 이 기술이 νΌμ§‘λ‹ˆλ‹€.
03:17
This is just representative, the chest X-ray.
54
197652
2628
ν•œ 예둜 흉뢀 μ—‘μŠ€λ ˆμ΄μž…λ‹ˆλ‹€.
03:20
And in fact with the chest X-ray,
55
200822
2169
μ‹€μ œλ‘œ 흉뢀 μ—‘μŠ€λ ˆμ΄λ₯Ό κ°€μ§€κ³ 
03:22
the ability here for the AI to pick up,
56
202991
3086
AI κ°€ λ­”κ°€λ₯Ό λ°œκ²¬ν•˜λŠ” λŠ₯λ ₯이 μ–΄λŠ 정도냐 ν•˜λ©΄,
03:26
the radiologists, expert radiologists missing the nodule,
57
206077
4880
μ „λ¬Έ 방사선사가 λ†“μΉœ κ²°μ ˆμ„
03:30
which turned out to be picked up by the AI as cancerous,
58
210999
3128
AIκ°€ 악성이라고 μ§‘μ–΄λƒˆμŠ΅λ‹ˆλ‹€.
03:34
and this is, of course, representative of all of medical scans,
59
214127
4004
λ¬Όλ‘  이것은 CTλ“ , MRIλ“ , μ΄ˆμŒνŒŒλ“ 
03:38
whether it’s CT scans, MRI, ultrasound.
60
218173
3837
λͺ¨λ“  의료 μ˜μƒμ— μ μš©λ©λ‹ˆλ‹€.
ν‘œμ§€κ°€ λΆ™κ³  주석이 달린 λŒ€κ·œλͺ¨ λ°μ΄ν„°λ‘œ 지도 ν•™μŠ΅μ„ ν•˜λ©΄
03:42
That through supervised learning of large, labeled, annotated data sets,
61
222051
5881
03:47
we can see AI do at least as well, if not better,
62
227974
3921
AIλŠ” μ „λ¬Έ μ˜μ‚¬λ³΄λ‹€ 더 λ‚«κ±°λ‚˜ μ΅œμ†Œν•œ 그에 λͺ»μ§€μ•ŠμŠ΅λ‹ˆλ‹€.
03:51
than expert physicians.
63
231895
1960
κ²°μž₯ λ‚΄μ‹œκ²½μ— λ¨Έμ‹  비전을 μ μš©ν•΄
03:55
And 21 randomized trials of picking up polyps --
64
235023
4755
μš©μ’…μ„ μ°Ύμ•„λ‚΄λŠ” λ¬΄μž‘μœ„ μ‹€ν—˜μ„ 21번 ν–ˆλŠ”λ°
03:59
machine vision during colonoscopy -- have all shown
65
239819
4171
λͺ¨λ“  μ‹€ν—˜μ—μ„œ μ†Œν™”κΈ° μ „λ¬Έμ˜ λ‹¨λ…μœΌλ‘œ ν•  λ•Œλ³΄λ‹€
04:03
that polyps are picked up better
66
243990
3003
04:06
with the aid of machine vision than by the gastroenterologist alone,
67
246993
3796
λ¨Έμ‹  λΉ„μ „μ˜ 도움을 받을 λ•Œ μš©μ’…μ„ 더 잘 μ°Ύμ•„λƒˆμŠ΅λ‹ˆλ‹€.
04:10
especially as the day goes on, later in the day, interestingly.
68
250789
4337
특히 μ‹œκ°„μ΄ μ§€λ‚ μˆ˜λ‘, μž¬λ―Έμžˆκ²Œλ„ μ˜€ν›„κ°€ 될수둝 말이죠.
04:15
We don't know whether picking up all these additional polyps
69
255168
3253
μš©μ’…μ„ μΆ”κ°€λ‘œ λ°œκ²¬ν•˜λŠ” 것이
μ•”μ˜ μžμ—° κ²½κ³Όλ₯Ό λ°”κΎΈλŠ”μ§€λŠ” μ•Œ 수 μ—†μ§€λ§Œ
04:18
changes the natural history of cancers,
70
258463
2085
04:20
but it tells you about machine eyes,
71
260590
3086
기계 μ‹œκ°μ˜ λŠ₯λ ₯은 μ•Œλ €μ€λ‹ˆλ‹€.
04:23
the power of machine eyes.
72
263718
1376
04:25
Now that was interesting.
73
265470
2377
ν₯미둜운 κ²°κ³Όμž…λ‹ˆλ‹€.
04:27
But now still with deep learning models, not transformer models,
74
267889
5714
트랜슀포머 λͺ¨λΈμ΄ μ•„λ‹ˆλΌ λ”₯ λŸ¬λ‹ λͺ¨λΈλ§Œ 가지고도
04:33
we've seen and learned that the ability
75
273645
3253
컴퓨터 μ‹œκ°μ€ μ‚¬λžŒ 눈이 보지 λͺ»ν•œ 것을 λ°œκ²¬ν•˜λŠ” λŠ₯λ ₯이
04:36
for computer vision to pick up things that human eyes can't see
76
276898
5589
μƒλ‹Ήνžˆ λ›°μ–΄λ‚˜λ‹€λŠ” 사싀을 이제 보고 λ°°μ› μŠ΅λ‹ˆλ‹€.
04:42
is quite remarkable.
77
282487
1460
04:43
Here's the retina.
78
283988
1418
이것은 λ§λ§‰μž…λ‹ˆλ‹€.
04:46
Picking up the control of diabetes and blood pressure.
79
286074
3378
ν˜ˆλ‹Ήκ³Ό ν˜ˆμ•• 쑰절 λŠ₯λ ₯을 μ•Œ 수 있고
04:50
Kidney disease.
80
290495
1710
μ‹ μž₯ μ§ˆν™˜,
04:52
Liver and gallbladder disease.
81
292872
2586
κ°„ 및 λ‹΄λ‚­ μ§ˆν™˜,
04:56
The heart calcium score,
82
296084
2043
심μž₯ 칼슘 μˆ˜μΉ˜λ„ μ•Œμ•„λ‚΄λŠ”λ°
04:58
which you would normally get through a scan of the heart.
83
298127
4004
보톡은 심μž₯ 검사λ₯Ό ν•΄μ•Ό μ•Œ 수 μžˆμŠ΅λ‹ˆλ‹€.
μ–΄λ–€ μž„μƒ 증상도 λ‚˜νƒ€λ‚˜κΈ° 전에 μ•ŒμΈ ν•˜μ΄λ¨Έλ₯Ό μ°Ύμ•„λ‚΄κ³ 
05:03
Alzheimer's disease before any clinical symptoms have been manifest.
84
303174
4129
심μž₯ λ§ˆλΉ„μ™€ λ‡Œμ‘Έμ€‘λ„ μ˜ˆμΈ‘ν•˜λ©°
05:08
Predicting heart attacks and strokes.
85
308012
2586
05:11
Hyperlipidemia.
86
311599
1543
κ³ μ§€ν˜ˆμ¦,
05:13
And seven years before any symptoms of Parkinson's disease,
87
313518
4546
증상이 λ‚˜νƒ€λ‚˜κΈ° 7λ…„ 전에 νŒŒν‚¨μŠ¨λ³‘λ„ μ°Ύμ•„λƒ…λ‹ˆλ‹€.
05:18
to pick that up.
88
318064
1251
05:19
Now this is interesting because in the future,
89
319941
3587
λŒ€λ‹¨νžˆ ν₯λ―Έλ‘­μŠ΅λ‹ˆλ‹€.
λ―Έλž˜μ—λŠ” κ²€μ§„ν•  λ•Œ 망막 사진을 찍게 될 ν…Œλ‹ˆκΉŒμš”.
05:23
we'll be taking pictures of our retina at checkups.
90
323570
3753
05:27
This is the gateway to almost every system in the body.
91
327365
3462
μ—¬κΈ°κ°€ μ‹ μ²΄μ˜ 거의 λͺ¨λ“  κΈ°κ΄€μœΌλ‘œ ν†΅ν•˜λŠ” 관문인 κ²λ‹ˆλ‹€.
05:31
It's really striking.
92
331369
1168
정말 λ†€λžμŠ΅λ‹ˆλ‹€.
05:32
And we'll come back to this because each one of these studies
93
332579
4087
이 μ—°κ΅¬λŠ” λ‚˜μ€‘μ— λ‹€μ‹œ λ³΄κ² μŠ΅λ‹ˆλ‹€.
05:36
was done with tens or hundreds [of] thousands of images
94
336666
4213
각각이 μˆ˜λ§Œμ—μ„œ μˆ˜μ‹­λ§Œ 개 μ˜μƒμœΌλ‘œ 지도 ν•™μŠ΅μ„ ν•œ 산물이고
05:40
with supervised learning,
95
340879
1251
05:42
and they’re all separate studies by different investigators.
96
342171
3921
μ „λΆ€ λ‹€λ₯Έ μ—°κ΅¬μžλ“€μ΄ μˆ˜ν–‰ν•œ λ³„κ°œ μ—°κ΅¬κ±°λ“ μš”.
05:46
Now, as a cardiologist, I love to read cardiograms.
97
346426
4045
μ €λŠ” 심μž₯병 μ „λ¬Έμ˜λ‘œμ„œ
심전도λ₯Ό μ½λŠ” 것을 μ’‹μ•„ν•˜κ³  이 일을 30λ…„ λ„˜κ²Œ ν•΄ μ™”μŠ΅λ‹ˆλ‹€.
05:50
I've been doing it for over 30 years.
98
350513
2169
05:53
But I couldn't see these things.
99
353808
2086
그런데 이런 것듀은 μ•Œ μˆ˜κ°€ μ—†μ—ˆμŠ΅λ‹ˆλ‹€.
05:56
Like, the age and the sex of the patient,
100
356519
2920
ν™˜μžμ˜ λ‚˜μ΄μ™€ 성별에
05:59
or the ejection fraction of the heart,
101
359439
3086
심μž₯의 λ°•μΆœλ₯ λ„ μ•Œμ•„λ‚΄κ³ 
06:02
making difficult diagnoses that are frequently missed.
102
362567
3503
ν”νžˆ μ‰½κ²Œ λ†“μΉ˜λŠ” μ–΄λ €μš΄ 진단도 λ‚΄λ¦½λ‹ˆλ‹€.
06:06
The anemia of the patient, that is, the hemoglobin to the decimal point.
103
366571
4212
μ†Œμˆ˜μ  λ‹¨μœ„λ‘œ ν˜ˆμƒ‰μ†Œ μˆ˜μΉ˜μ™€ 빈혈 여뢀도 μ•Œ 수 있고
06:11
Predicting whether a person,
104
371951
1460
μ‹¬λ°©μ„Έλ™μ΄λ‚˜ λ‡Œμ‘Έμ€‘μ„ κ²ͺ은 적 μ—†λŠ” μ‚¬λžŒμ—κ²Œ,
06:13
who's never had atrial fibrillation or stroke
105
373453
2502
06:15
from the ECG,
106
375955
1418
그런 병이 생길 κ°€λŠ₯성이 μžˆλŠ”μ§€λ„ μ‹¬μ „λ„λ‘œ μ˜ˆμΈ‘ν•©λ‹ˆλ‹€.
06:17
whether that's going to likely occur.
107
377415
2169
06:20
Diabetes, a diagnosis of diabetes and prediabetes, from the cardiogram.
108
380418
4796
당뇨병과 당뇨병 전증 여뢀도 μ‹¬μ „λ„λ‘œ μ°Ύμ•„λƒ…λ‹ˆλ‹€.
06:25
The filling pressure of the heart.
109
385965
2044
심μž₯의 좩만기 μ••λ ₯,
06:28
Hypothyroidism
110
388509
2086
갑상선 κΈ°λŠ₯ 항진증 μ—¬λΆ€,
06:30
and kidney disease.
111
390637
1626
μ‹ μž₯ μ§ˆν™˜.
06:32
Imagine getting an electrocardiogram to tell you about all these other things,
112
392305
3920
μ΄λ ‡κ²Œ 심μž₯κ³Ό κ΄€κ³„μ—†λŠ” λͺ¨λ“  것듀에 λŒ€ν•΄
μ‹¬μ „λ„λ§ŒμœΌλ‘œ μ•Œ 수 μžˆλ‹€κ³  상상해 λ³΄μ„Έμš”.
06:36
not really so much about the heart.
113
396267
2711
06:39
Then there's the chest X-ray.
114
399729
1543
흉뢀 μ—‘μŠ€λ ˆμ΄λ„ μžˆμŠ΅λ‹ˆλ‹€.
06:41
Who would have guessed that we could accurately determine
115
401314
3920
ν™˜μžμ˜ 인쒅을 μ •ν™•νžˆ νŒλ³„ν•  수 μžˆλ‹€κ³  λˆ„κ°€ μƒμƒμ΄λ‚˜ ν–ˆμ„κΉŒμš”?
06:45
the race of the patient,
116
405234
1377
06:46
no less the ethical implications of that,
117
406611
2794
μ΄κ²ƒμ˜ 윀리적 μ˜λ―ΈλŠ” 말할 ν•„μš”λ„ μ—†κ² μ§€λ§Œ,
06:49
from a chest X-ray through machine eyes?
118
409405
3379
기계 눈으둜 흉뢀 μ—‘μŠ€λ ˆμ΄λ₯Ό 보고 νŒλ³„ν•  수 μžˆμŠ΅λ‹ˆλ‹€.
06:53
And interestingly, picking up the diagnosis of diabetes,
119
413201
4171
그리고 ν₯λ―Έλ‘­κ²Œλ„ 당뇨병 여뢀와
06:57
as well as how well the diabetes is controlled,
120
417372
4212
당뇨가 μ–Όλ§ˆλ‚˜ 잘 κ΄€λ¦¬λ˜κ³  μžˆλŠ”μ§€λ„
07:01
through the chest X-ray.
121
421584
1668
흉뢀 μ—‘μŠ€λ ˆμ΄λ‘œ νŒλ‹¨ν•©λ‹ˆλ‹€.
07:04
And of course, so many different parameters about the heart,
122
424629
3795
λ¬Όλ‘  심μž₯에 λŒ€ν•œ 맀개 λ³€μˆ˜κ°€ λ„ˆλ¬΄ 많고 λ‹€μ–‘ν•΄μ„œ
07:08
which we could never,
123
428424
2169
우리 방사선 μ „λ¬Έμ˜, 심μž₯병 μ „λ¬Έμ˜λŠ” μ ˆλŒ€λ‘œ
07:10
radiologists or cardiologists, could never be able to come up
124
430593
3837
λ¨Έμ‹  비전이 ν•  수 μžˆλŠ” 이런 일듀을 생각해낼 수 μ—†μ—ˆμŠ΅λ‹ˆλ‹€.
07:14
with what machine vision can do.
125
434430
2878
07:17
Pathologists often argue about a slide,
126
437976
3169
λ³‘λ¦¬ν•™μžλ“€μ€ ν”νžˆ μŠ¬λΌμ΄λ“œ ν•˜λ‚˜λ₯Ό 두고
07:21
about what does it really show?
127
441187
1794
그게 뭘 λ‚˜νƒ€λ‚΄λŠ”μ§€ λ…ΌμŸμ„ λ²Œμž…λ‹ˆλ‹€.
07:23
But with this ability of machine eyes,
128
443314
4338
ν•˜μ§€λ§Œ 이 기계 눈의 λŠ₯λ ₯으둜
07:27
the driver genomic mutations of the cancer can be defined,
129
447694
3878
암을 μœ λ°œν•˜λŠ” μ£Όμš” μœ μ „μ²΄ 변이뿐 μ•„λ‹ˆλΌ
07:31
no less the structural copy number variants
130
451614
2878
쒅양에 λ‚˜νƒ€λ‚˜λŠ” ꡬ쑰적인 볡제수 변이도 μ •μ˜ν•  수 μžˆμŠ΅λ‹ˆλ‹€.
07:34
that are accounting or present in that tumor.
131
454534
2878
07:37
Also, where is that tumor coming from?
132
457787
2336
또, λ§Žμ€ ν™˜μžλ“€μ—μ„œλŠ” 쒅양이 μ–΄λ””μ—μ„œ λ²ˆμ‘ŒλŠ”μ§€ μ•Œ 수 μ—†μŠ΅λ‹ˆλ‹€.
07:40
For many patients, we don’t know.
133
460164
2253
07:42
But it can be determined through AI.
134
462458
4255
ν•˜μ§€λ§Œ AIλ₯Ό 톡해 확인할 수 μžˆμŠ΅λ‹ˆλ‹€.
07:46
And also the prognosis of the patient,
135
466754
2836
λ˜ν•œ ν™˜μžμ˜ μ˜ˆν›„ μ—­μ‹œ μŠ¬λΌμ΄λ“œλ§ŒμœΌλ‘œ μ•Œ 수 μžˆμŠ΅λ‹ˆλ‹€.
07:49
just from the slide,
136
469590
2169
07:51
by all of the training.
137
471801
1627
이런 λͺ¨λ“  ν•™μŠ΅μ„ ν•˜λ©΄ 말이죠.
07:53
Again, this is all just convolutional neural networks,
138
473469
4797
λ‹€μ‹œ λ§μ”€λ“œλ¦¬μ§€λ§Œ, μ—¬κΈ°κΉŒμ§„ ν•©μ„±κ³± 신경망일 뿐이며
07:58
not transformer models.
139
478307
1669
트랜슀포머 λͺ¨λΈμ€ μ•„λ‹™λ‹ˆλ‹€.
08:00
So when we go from the deep neural networks to transformer models,
140
480852
5630
이 심측 μ‹ κ²½λ§μ—μ„œ 트랜슀포머 λͺ¨λΈλ‘œ λ„˜μ–΄κ°€ λ΄…μ‹œλ‹€.
08:06
this classic pre-print,
141
486524
2085
이 고전적인 μ˜ˆλΉ„ 논문은
08:08
one of the most cited pre-prints ever,
142
488651
2586
μ—­λŒ€ 무척 많이 인용된 μ˜ˆλΉ„ λ…Όλ¬Έ 쀑 ν•˜λ‚˜μΈ
08:11
"Attention is All You Need,"
143
491237
1418
'Attention is All You Need'μž…λ‹ˆλ‹€.
08:12
the ability to now be able to look at many more items,
144
492697
4296
이 κΈ°μˆ μ€ μ–Έμ–΄λ“  μ˜μƒμ΄λ“ 
더 λ§Žμ€ μš”μ†Œλ₯Ό 보고
08:17
whether it be language or images,
145
497035
3837
08:20
and be able to put this in context,
146
500913
2962
λ§₯락에 맞좰 κ΅¬μ„±ν•΄μ„œ
08:23
setting up a transformational progress in many fields.
147
503916
4588
λ§Žμ€ λΆ„μ•Όμ—μ„œ ν˜μ‹ μ μΈ 진전을 μ΄λ€˜μŠ΅λ‹ˆλ‹€.
08:29
The prototype is, the outgrowth of this is GPT-4.
148
509172
4504
μ‹œν—˜ μž‘ν’ˆμ€ μ§€ν”Όν‹°4μž…λ‹ˆλ‹€.
08:34
With over a trillion connections.
149
514510
2628
μ§€ν”Όν‹°4μ—λŠ” 1μ‘° κ°œκ°€ λ„˜λŠ” 연결이 μžˆμŠ΅λ‹ˆλ‹€.
08:37
Our human brain has 100 trillion connections or parameters.
150
517138
4713
μΈκ°„μ˜ λ‡Œμ—λŠ” μ—°κ²° λ˜λŠ” 맀개 λ³€μˆ˜κ°€ 100μ‘° 개 μžˆμŠ΅λ‹ˆλ‹€.
08:42
But one trillion,
151
522185
1167
ν•˜μ§€λ§Œ 1μ‘° 개 연결에 λ“€μ–΄κ°„ 정보와 지식을 상상해 λ³΄μ„Έμš”.
08:43
just think of all the information, knowledge,
152
523352
2128
08:45
that's packed into those one trillion.
153
525480
1876
08:47
And interestingly, this is now multimodal with language, with images,
154
527398
4880
μ΄μ œλŠ” 거기에 문자, μ˜μƒ, μŒμ„± λ“± 볡합적인 데이터가 μžˆμ–΄μ„œ
08:52
with speech.
155
532320
1376
08:53
And it involves a massive amount of graphic processing units.
156
533696
3921
κ·Έλž˜ν”½ 처리 μž₯μΉ˜κ°€ μ—„μ²­λ‚˜κ²Œ 많이 ν•„μš”ν•©λ‹ˆλ‹€.
그리고 자기 지도 ν•™μŠ΅μ„ ν•©λ‹ˆλ‹€.
08:58
And it's with self-supervised learning,
157
538076
2293
09:00
which is a big bottleneck in medicine
158
540369
2128
μ„€λͺ… λΆ™μ΄κΈ°λŠ” μ˜ν•™μ—μ„œ 큰 병λͺ©μΈλ°
09:02
because we can't get experts to label images.
159
542497
3169
μ˜μƒμ— μ„€λͺ…을 뢙일 μ „λ¬Έκ°€λ₯Ό κ΅¬ν•˜κΈ° νž˜λ“€κ±°λ“ μš”.
09:05
This can be done with self-supervised learning.
160
545708
2795
이 문제λ₯Ό 자기 지도 ν•™μŠ΅μ„ 톡해 ν•΄κ²°ν•  수 있죠.
09:08
So what does this set up in medicine?
161
548961
2837
κ·Έλ ‡λ‹€λ©΄ 이것이 μ˜ν•™μ—λŠ” μ–΄λ–€ 영ν–₯을 λ―ΈμΉ κΉŒμš”?
09:11
It sets up, for example, keyboard liberation.
162
551839
4421
예λ₯Ό λ“€μ–΄ μžνŒμ—μ„œ 해방될 수 μžˆμŠ΅λ‹ˆλ‹€.
09:16
The one thing that both doctors, clinicians
163
556803
3920
μ˜μ‚¬, μž„μƒμ˜, ν™˜μž λͺ¨λ‘κ°€ μ›ν•˜λŠ” κ²ƒμž…λ‹ˆλ‹€.
09:20
and patients would like to see.
164
560765
2377
09:23
Everyone hates being data clerks as clinicians,
165
563851
3921
데이터 μž…λ ₯은 λˆ„κ΅¬λ‚˜ μ‹«μ–΄ν•©λ‹ˆλ‹€.
ν™˜μžλ“€μ€ 였래 κΈ°λ‹€λ¦° 끝에 λ“œλ””μ–΄ 병원에 였면
09:27
and patients would like to see their doctor
166
567814
2836
09:30
when they finally have the visit they've waited for a long time.
167
570650
3753
μ˜μ‚¬λ₯Ό 보고 μ‹Άμ–΄ν•©λ‹ˆλ‹€.
09:34
So the ability to change the face-to-face contact
168
574445
4713
λŒ€λ©΄ μ§„λ£Œμ˜ κ²½ν—˜μ„ λ°”κΎΈλŠ” 것은
이제 λ°”λ‘œ ν•œ 발짝 μ•žμ— 와 μžˆμŠ΅λ‹ˆλ‹€.
09:39
is just one step along the way.
169
579200
2502
09:41
By having the liberation from keyboards with synthetic notes
170
581744
5005
λŒ€ν™”μ—μ„œ λ„μΆœλœ λ‚΄μš©μœΌλ‘œ λ…ΈνŠΈλ₯Ό λ§Œλ“€μ–΄μ„œ
09:46
that are driven, derived from the conversation,
171
586791
2753
μžνŒμ—μ„œ ν•΄λ°©λœλ‹€λ©΄
09:49
and then all the downstream normal data clerk functions that are done,
172
589585
4880
일반적인 후속 μž‘μ—…μΈ 데이터λ₯Ό μž…λ ₯ν•˜λŠ” 뢀뢄은
09:54
often off-hours.
173
594507
1668
λŒ€κ°œ 업무가 λλ‚˜κ³  μ²˜λ¦¬λ˜λŠ” κ±°μ£ .
09:56
Now we're seeing in health systems across the United States
174
596217
3587
이제 μš°λ¦¬λŠ” λ―Έκ΅­ μ „μ—­μ˜ 보건 μ²΄κ³„μ—μ„œ
09:59
where people, physicians are saving many hours of time
175
599846
3920
μ‚¬λžŒλ“€, μ˜μ‚¬λ“€μ΄ λ§Žμ€ μ‹œκ°„μ„ μ ˆμ•½ν•˜κ³ ,
10:03
and heading towards ultimately keyboard liberation.
176
603808
3587
ꢁ극적으둜 자판이 μ—†λŠ” 곳을 ν–₯ν•΄ κ°€λŠ” κ±Έ 보고 μžˆμŠ΅λ‹ˆλ‹€.
10:08
We recently published, with the group at Moorfields Eye Institute,
177
608396
3587
μš°λ¦¬λŠ” μ΅œκ·Όμ— ν”Όμ–΄μŠ€ 킨이 μ΄λ„λŠ” λ¬΄μ–΄ν•„λ“œ μ•ˆκ³Ό μ—°κ΅¬μ†Œ νŒ€κ³Ό ν•¨κ»˜
10:12
led by Pearse Keane,
178
612024
1335
10:13
the first foundation model in medicine from the retina.
179
613401
3295
망막 μ§„λ‹¨μš© 기초 λͺ¨λΈμ„ 처음 λ‚΄λ†“μ•˜μŠ΅λ‹ˆλ‹€.
10:16
And remember those eight different things that were all done by separate studies?
180
616737
4380
각각 λ‹€λ₯Έ μ—°κ΅¬μžκ°€ ν•œ μ—¬λŸ κ°€μ§€ 연ꡬ가 κΈ°μ–΅λ‚˜μ‹œλ‚˜μš”?
10:21
This was all done with one model.
181
621159
2335
그게 λ‹€ ν•œ λͺ¨λΈλ‘œ μ΄λ£¨μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€.
10:23
This is with 1.6 million retinal images
182
623494
3879
망막 μ˜μƒ 160만 개λ₯Ό 톡해
10:27
predicting all these different outcome likelihoods.
183
627415
4546
μ΄λ ‡κ²Œ λ‹€μ–‘ν•œ κ²°λ‘ λ“€μ˜ κ°€λŠ₯성을 μ˜ˆμΈ‘ν•  수 μžˆμŠ΅λ‹ˆλ‹€.
κ²Œλ‹€κ°€ λͺ¨λ‘ κ³΅κ°œλ˜μ–΄ μžˆλŠ”λ°
10:32
And this is all open-source,
184
632003
1710
10:33
which is of course really important that others can build on these models.
185
633754
4380
λ‹€λ₯Έ μ‚¬λžŒλ“€μ΄ 이 λͺ¨λΈμ„ 기초둜 μ‚Όμ•„ κ°œλ°œν•  수 μžˆμ–΄μ„œ 정말 μ€‘μš”ν•©λ‹ˆλ‹€.
10:38
Now I just want to review a couple of really interesting patients.
186
638134
5547
이제 정말 ν₯미둜운 ν™˜μž λͺ‡ λͺ…에 λŒ€ν•΄ 이야기 ν•΄ λ³΄κ² μŠ΅λ‹ˆλ‹€.
μ•€λ“œλ₯˜λŠ” 이제 μ—¬μ„― μ‚΄μž…λ‹ˆλ‹€.
10:44
Andrew, who is now six years old.
187
644098
3003
10:47
He had three years of relentlessly increasing pain, arrested growth.
188
647810
7007
3λ…„ λ™μ•ˆ λŠμž„μ—†λŠ” 고톡이 점점 μ‹¬ν•΄μ‘Œκ³ 
μ„±μž₯이 λ©ˆμ·„μŠ΅λ‹ˆλ‹€.
10:55
His gait suffered with a dragging of his left foot,
189
655318
2544
μ™Όλ°œμ΄ 질질 끌렀 κ±·κΈ°κ°€ νž˜λ“€μ—ˆμœΌλ©° μ‹¬ν•œ 두톡에 μ‹œλ‹¬λ ΈμŠ΅λ‹ˆλ‹€.
10:57
he had severe headaches.
190
657862
1918
10:59
He went to 17 doctors over three years.
191
659780
3337
μ•€λ“œλ₯˜λŠ” 3λ…„ λ™μ•ˆ μ˜μ‚¬ 17λͺ…을 μ°Ύμ•„κ°”μŠ΅λ‹ˆλ‹€.
11:03
His mother then entered all his symptoms into ChatGPT.
192
663743
4254
κ·Έ ν›„ μ–΄λ¨Έλ‹ˆλŠ” μ•„μ΄μ˜ 증상을 챗지피티에 λͺ¨λ‘ μž…λ ₯ν–ˆμŠ΅λ‹ˆλ‹€.
11:08
It made the diagnosis of occulta spina bifida,
193
668706
4254
κ±°κΈ°μ„œ μž μž¬μ„± 이뢄 μ²™μΆ”λΌλŠ” 진단을 μ–»μ—ˆμŠ΅λ‹ˆλ‹€.
11:12
which meant he had a tethered spinal cord that was missed by all 17 doctors
194
672960
5297
κ·Έ 말은 μ²™μˆ˜κ°€ λ¬Άμ—¬ μžˆλŠ” μƒνƒœμ˜€λŠ”λ°
3λ…„ λ™μ•ˆ μ˜μ‚¬ 17λͺ…이 λͺ¨λ‘ λ†“μ³€λ‹€λŠ” κ²λ‹ˆλ‹€.
11:18
over three years.
195
678257
1168
11:19
He had surgery to release the cord.
196
679467
2002
μ²™μˆ˜λ₯Ό ν’€μ–΄μ£ΌλŠ” μˆ˜μˆ μ„ 받은 λ’€ μ•„μ΄λŠ” 이제 μ™„λ²½νžˆ κ±΄κ°•ν•©λ‹ˆλ‹€.
11:21
He's now perfectly healthy.
197
681469
1793
11:24
(Applause)
198
684889
5630
(λ°•μˆ˜)
11:30
This is a patient that was sent to me,
199
690561
2920
μ œκ°€ μ§„λ£Œ 의뒰λ₯Ό 받은 μ–΄λ–€ ν™˜μžλŠ”
11:33
who was suffering with, she was told, long COVID.
200
693481
4671
μž₯κΈ° μ½”λ‘œλ‚˜λΌλŠ” 진단을 λ°›κ³  고톡을 λ°›κ³  μžˆμ—ˆμŠ΅λ‹ˆλ‹€.
11:38
She saw many different physicians, neurologists,
201
698694
3379
이 ν™˜μžλŠ” μ—¬λŸ¬ μ˜μ‚¬μ™€ μ‹ κ²½μ™Έκ³Ό μ „λ¬Έμ˜λ₯Ό λ§Œλ‚¬λŠ”λ°
11:42
and her sister entered all her symptoms after getting nowhere,
202
702073
4546
μ–΄λ–€ 진전도 μ—†μž κ·Έλ…€μ˜ 여동생이 λͺ¨λ“  증상을 μž…λ ₯ν–ˆμŠ΅λ‹ˆλ‹€.
11:46
no treatment for long COVID,
203
706619
1418
μž₯κΈ° μ½”λ‘œλ‚˜ 치료λ₯Ό λ°›μ§€ λͺ»ν–ˆκ³  κ²€μ¦λœ μΉ˜λ£Œλ²•λ„ μ—†μŠ΅λ‹ˆλ‹€.
11:48
there is no treatment validated,
204
708079
1710
11:49
and her sister put all her symptoms into ChatGPT.
205
709789
4421
κ·Έλž˜μ„œ ν™˜μžμ˜ 여동생이 λͺ¨λ“  증상을 챗지피티에 μž…λ ₯ν•΄μ„œ
11:54
It found out it actually was not long COVID,
206
714252
2293
사싀은 μž₯κΈ° μ½”λ‘œλ‚˜κ°€ μ•„λ‹ˆλΌ
11:56
she had limbic encephalitis, which is treatable.
207
716587
3462
변연계 λ‡Œμ—Όμ΄λž€ κ±Έ μ•Œμ•„λƒˆκ³  이건 μΉ˜λ£Œκ°€ κ°€λŠ₯ν•œ κ²ƒμž…λ‹ˆλ‹€.
12:00
She was treated, and now she's doing extremely well.
208
720091
3128
이 ν™˜μžλŠ” 치료λ₯Ό λ°›μ•˜κ³  μ§€κΈˆμ€ μƒνƒœκ°€ μ•„μ£Ό μ’‹μŠ΅λ‹ˆλ‹€.
12:03
But these are not just anecdotes anymore.
209
723594
2753
그런데 이것듀은 더 이상 λ‹¨μˆœν•œ 일화가 μ•„λ‹™λ‹ˆλ‹€.
12:06
70 very difficult cases
210
726389
3461
μ•„μ£Ό μ–΄λ €μš΄ 진단 사둀 70건을
12:09
that are the clinical pathologic conferences
211
729850
2461
뉴 μž‰κΈ€λžœλ“œ μ˜ν•™ μ €λ„μ˜ μž„μƒλ³‘λ¦¬ν•™νšŒμ—μ„œ 골라
12:12
at the New England Journal of Medicine
212
732353
1877
12:14
were compared to GPT-4,
213
734272
2836
μ§€ν”Όν‹°4와 λΉ„κ΅ν•΄λ΄€μŠ΅λ‹ˆλ‹€.
12:17
and the chatbot did as well
214
737149
3295
챗봇은 진단을 λ‚΄λ¦¬λŠ” 데에 μ „λ¬Έ μž„μƒμ˜μ™€ κ°™κ±°λ‚˜
12:20
or better than the expert master clinicians
215
740486
3295
더 λ‚˜μ€ λͺ¨μŠ΅μ„ λ³΄μ—¬μ€¬μŠ΅λ‹ˆλ‹€.
12:23
in making the diagnosis.
216
743781
1960
12:26
So I just want to close with a recent conversation with my fellow.
217
746492
4713
μ΅œκ·Όμ— λ™λ£Œμ™€ λ‚˜λˆˆ λŒ€ν™”λ‘œ 마무리λ₯Ό μ§“κ² μŠ΅λ‹ˆλ‹€.
12:31
Medicine is still an apprenticeship,
218
751706
2085
μ˜ν•™μ€ μ—¬μ „νžˆ κ²¬μŠ΅μƒ μ œλ„λ‘œ κ΅μœ‘ν•˜λ©°
12:33
and Andrew Cho is 30 years old,
219
753833
3837
μ•€λ“œλ₯˜ μ‘°λŠ” μ„œλ₯Έ 살인데
12:37
in his second year of cardiology fellowship.
220
757670
2085
심μž₯ν•™ μ „λ¬Έμ˜ 이 λ…„μ°¨μž…λ‹ˆλ‹€.
12:39
We see all patients together in the clinic.
221
759797
2669
μš°λ¦¬λŠ” μ§„λ£Œμ‹€μ—μ„œ λͺ¨λ“  ν™˜μžλ₯Ό ν•¨κ»˜ λ΄…λ‹ˆλ‹€.
12:42
And at the end of clinic the other day,
222
762967
2252
ν•˜λ£¨λŠ” μ§„λ£Œκ°€ 끝날 무렡에
12:45
I sat down and said to him,
223
765261
1918
μ œκ°€ μžλ¦¬μ— μ•‰μœΌλ©° λ§ν–ˆμ£ .
12:47
"Andrew, you are so lucky.
224
767179
2795
β€œμ•€λ“œλ₯˜, 당신은 정말 운이 μ’‹μ•„μš”."
12:50
You're going to be practicing medicine in an era of keyboard liberation.
225
770516
4838
"자판이 ν•„μš” μ—†λŠ” μ‹œλŒ€μ— μ§„λ£Œλ₯Ό ν•˜κ²Œ 될 κ±°μ˜ˆμš”."
12:55
You're going to be connecting with patients
226
775813
2044
"μˆ˜μ‹­ λ…„κ°„ μš°λ¦¬κ°€ ν•΄ 보지 μ•Šμ€ λ°©μ‹μœΌλ‘œ ν™˜μžμ™€ μ†Œν†΅ν•˜κ²Œ λ˜λŠ” κ±°μ£ ."
12:57
the way we haven't done for decades."
227
777857
2502
13:00
That is the ability to have the note
228
780735
3086
이 기술이 λŒ€ν™”λ₯Ό λ“£κ³  λ…ΈνŠΈλ„ 써 μ£Όκ³ 
13:03
and the work from the conversation
229
783863
2502
λŒ€ν™”λ₯Ό λ“£κ³  일도 λŒ€μ‹ ν•΄ μ£ΌλŠ” κ²λ‹ˆλ‹€.
13:06
to derive things like pre-authorization,
230
786407
3795
예λ₯Ό λ“€μ–΄ 사전 μŠΉμΈμ΄λ‚˜
13:10
billing, prescriptions, future appointments --
231
790202
4755
μ²­κ΅¬μ„œλ₯Ό μž‘μ„±ν•˜κ³ , μ²˜λ°©ν•˜κ³ , λ‹€μŒ μ§„λ£Œλ₯Ό μ˜ˆμ•½ν•˜λŠ” κ²ƒκΉŒμ§€μš”.
13:14
all the things that we do,
232
794999
1293
ν™˜μžλ“€μ„ μœ λ„ν•˜λŠ” κ²ƒκΉŒμ§€ μš°λ¦¬κ°€ ν•˜λŠ” λͺ¨λ“  일듀 λ§μž…λ‹ˆλ‹€.
13:16
including nudges to the patient.
233
796334
1584
13:17
For example, did you get your blood pressure checks
234
797918
2461
예λ₯Ό λ“€μ–΄ ν˜ˆμ•• 검사λ₯Ό λ°›μ•˜λŠ”μ§€ 묻고
13:20
and what did they show
235
800421
1168
κ²°κ³Όκ°€ μ–΄λ• λŠ”μ§€ 물으면 그것듀이 λͺ¨λ‘ μ—¬λŸ¬λΆ„μ—κ²Œ λŒμ•„μ˜΅λ‹ˆλ‹€.
13:21
and all that coming back to you.
236
801630
1544
13:23
But much more than that,
237
803215
1710
ν•˜μ§€λ§Œ 훨씬 μ€‘μš”ν•œ 것은 진단에 도움을 μ£ΌλŠ” κ²λ‹ˆλ‹€.
13:24
to help with making diagnoses.
238
804925
2086
13:27
And the gift of time
239
807928
2002
그리고 μ‹œκ°„μ„ λ²Œμ–΄ 주기도 ν•˜λŠ”λ°
13:29
that having all the data of a patient
240
809972
2169
ν™˜μžμ˜ λͺ¨λ“  정보λ₯Ό λ§Œλ‚˜κΈ°λ„ 전에 μ€€λΉ„ν•΄ μ€λ‹ˆλ‹€.
13:32
that's all teed up before even seeing the patient.
241
812183
2961
13:35
And all this support changes the future of the patient-doctor relationship,
242
815144
6632
이런 도움이 μ•žμœΌλ‘œ ν™˜μžμ™€ μ˜μ‚¬μ˜ 관계λ₯Ό λ°”κΏ”μ„œ
13:41
bringing in the gift of time.
243
821776
2460
μ‹œκ°„μ΄λž€ 선물을 κ°€μ Έλ‹€ μ€λ‹ˆλ‹€.
13:44
So this is really exciting.
244
824612
1710
정말 μ‹ λ‚˜λŠ” 일이죠.
13:46
I said to Andrew, everything has to be validated, of course,
245
826364
4295
μ €λŠ” μ•€λ“œλ₯˜μ—κ²Œ, λ¬Όλ‘  λͺ¨λ“  κ±Έ 검증해야 ν•œλ‹€κ³  ν–ˆμŠ΅λ‹ˆλ‹€.
13:50
that the benefit greatly outweighs any risk.
246
830701
3796
이읡이 μœ„ν—˜μ„±μ„ 훨씬 μƒνšŒν•œλ‹€λŠ” κ±Έ 확인해야겠죠.
13:54
But it is really a remarkable time for the future of health care,
247
834538
4505
κ·Έλž˜λ„ μ§€κΈˆμ€ 의료 μ„œλΉ„μŠ€μ˜ 미래λ₯Ό μœ„ν•œ 정말 λ†€λΌμš΄ μ‹œκΈ°μž…λ‹ˆλ‹€.
13:59
it's so damn exciting.
248
839085
2544
정말 λ„ˆλ¬΄λ‚˜λ„ μ‹ λ‚˜λŠ” μΌμž…λ‹ˆλ‹€.
14:01
Thank you.
249
841962
1168
κ³ λ§™μŠ΅λ‹ˆλ‹€.
(λ°•μˆ˜)
14:03
(Applause)
250
843172
2753
이 μ›Ήμ‚¬μ΄νŠΈ 정보

이 μ‚¬μ΄νŠΈλŠ” μ˜μ–΄ ν•™μŠ΅μ— μœ μš©ν•œ YouTube λ™μ˜μƒμ„ μ†Œκ°œν•©λ‹ˆλ‹€. μ „ 세계 졜고의 μ„ μƒλ‹˜λ“€μ΄ κ°€λ₯΄μΉ˜λŠ” μ˜μ–΄ μˆ˜μ—…μ„ 보게 될 κ²ƒμž…λ‹ˆλ‹€. 각 λ™μ˜μƒ νŽ˜μ΄μ§€μ— ν‘œμ‹œλ˜λŠ” μ˜μ–΄ μžλ§‰μ„ 더블 ν΄λ¦­ν•˜λ©΄ κ·Έκ³³μ—μ„œ λ™μ˜μƒμ΄ μž¬μƒλ©λ‹ˆλ‹€. λΉ„λ””μ˜€ μž¬μƒμ— 맞좰 μžλ§‰μ΄ μŠ€ν¬λ‘€λ©λ‹ˆλ‹€. μ˜κ²¬μ΄λ‚˜ μš”μ²­μ΄ μžˆλŠ” 경우 이 문의 양식을 μ‚¬μš©ν•˜μ—¬ λ¬Έμ˜ν•˜μ‹­μ‹œμ˜€.

https://forms.gle/WvT1wiN1qDtmnspy7