FILE: Beijing, China - Aug 26, 2017
1. Various of robot taking exam
2. Sign of National Medical Examination Center

Hefei City, Anhui Province, east China - Nov 14, 2017
3. Computer screen showing exam result
4. SOUNDBITE (Chinese) Wu Ji, deputy director, Electronic Engineering Department, Tsinghua University (starting with shot 3):
"Its score is 96 points above the acceptance line. This shows that it has indeed mastered the medical knowledge and clinical knowledge, and it has owned the basic ability to employ the knowledge to solve some problems."

FILE: Beijing, China - Aug 26, 2017
5. Various of robot taking exam
6. SOUNDBITE (Chinese) Invigilator, National Medical Examination Center (no name given):
"We made an electronic edition of the test paper for it, with the content same as the paper edition."
7. Robot taking exam
8. SOUNDBITE (Chinese) Wu Ji, deputy director, Electronic Engineering Department, Tsinghua University:
"We should say that robots are superior to people in memory, or storage capacity and computing ability."

FILE: Hefei City, Anhui Province, east China - Date Unknown
9. Researcher testing robots

FILE: Beijing, China - Aug 26, 2017
10. SOUNDBITE (Chinese) Wu Ji, deputy director, Electronic Engineering Department, Tsinghua University (partially overlaid with shot 11):
"Quite a few questions are case analyses, that is, you are informed the basic conditions of some patients and some of their symptoms, then you have to tell what disease it is, what treatment should be adopted or what medicine they should take. Many are questions like this. Such complicated questions cannot be answered only by searching. Therefore, it is essentially a reasoning process based on knowledge and probability model."

FILE: Hefei City, Anhui Province, east China - Date Unknown
++SHOT OVERLAYING SOUNDBITE++
11. Computer screen showing robot answering question
++SHOT OVERLAYING SOUNDBITE++
12. Various of robots

Hefei City, Anhui Province, east China - Nov 14, 2017
13. SOUNDBITE (Chinese) Wu Ji, deputy director, Electronic Engineering Department, Tsinghua University:
"What it can do most at present is to offer some possible suggestions to doctors, so as to help doctors to find where the problems lie more rapidly, or to avoid some risks."

FILE: Hefei City, Anhui Province, east China - Date Unknown
14. Various of robot working in hospital

A robot developed by Tsinghua University has passed China's National Medical Licensing Examination.

The robot named Xiao Yi scored 456, exceeding the acceptance line by 96 points. It is the first robot in China to participate in the exam with full marks of 600.

"Its score is 96 points above the acceptance line. This shows that it has indeed mastered the medical knowledge and clinical knowledge, and it has owned the basic ability to employ the knowledge to solve some problems," said Wu Ji, deputy director of electronic engineering department, Tsinghua University.

Xiao Yi took the exam in a separate room simultaneously with the other examinees, with nearly 10 staff members working for its exam.

"We made an electronic edition of the test paper for it, with the content same as the paper edition," said an invigilator of the National Medical Examination Center.

To get a good exam result, the researchers of Xiao Yi started to prepare for the exam since June 2016. They input knowledge from tens of medical textbooks into Xiao Yi's brain.

"We should say that robots are superior to people in memory, or storage capacity and computing ability," said Wu.

However, just remembering the knowledge is far from enough. Xiao Yi also had to analyze all the information provided by the test questions.

"Quite a few questions are case analyses, that is, you are informed the basic conditions of some patients and some of their symptoms, then you have to tell what disease it is, what treatment should be adopted or what medicine they should take. Many are questions like this. Such complicated questions cannot be answered only by searching. Therefore, it is essentially a reasoning process based on knowledge and probability model," said Wu.

To deal with the problem, researchers invited medical experts for Xiao Yi in February to learn clinical experience and case diagnosis. Yet Xiao Yi and its researchers still have a long way to go before becoming qualified for a doctor.

"What it can do most at present is to offer some possible suggestions to doctors, so as to help doctors to find where the problems lie more rapidly, or to avoid some risks," said Wu.