Aussie surgeon develops world-first program to determine liver transplant success likelihood before surgery
Source: Xinhua   2017-01-12 08:33:38

SYDNEY, Jan. 12 (Xinhua) -- An Australian surgeon has developed a world-first computer program to determine the likelihood of success of a liver transplant before the surgery takes place.

The program, developed by a junior surgeon at Melbourne's Austin Hospital, is inspired by the algorithm used by dating websites to pair two people together.

The algorithm uses the top 15 factors known at th time of the transplant to predict the early success of a liver donation and has proved significantly more successful than the current method which is the surgeon's intuition.

The world-first technology successfully predicted graft failure a month after surgery 84 percent of the time, compared to 68 percent for surgeons and researchers are optimistic that it could be adapted for other transplants.

Lawrence Lau, the developer of the program, said he was inspired to develop the machine-learning program by a friend who worked for popular dating website eHarmony.

"When we do the organ retrieval we look at the liver, we feel it with our hands, but ultimately a lot of clinical decisions are based on gut feeling," Lau told News Limited on Thursday.

"At the moment there is not a lot that guides the question of who you give the organ to, other than blood group and the urgency the transplant is required."

Using data from the University of Melbourne, the team from the Austin taught the program to learn using liver transplant outcomes from 2010 to 2013.

Lau said that he hoped the technique could offer thousands of Australians in need of a transplant a second chance at life with fewer risks.

"On one hand if we transplant an organ that is unsuitable, that recipient may die. But sadly, 10 percent of Australians waiting for a liver transplant die while waiting," he said.

"If we could quantify this risk, then we can start maximising the use of this precious resource."

Editor: Hou Qiang
Related News
Xinhuanet

Aussie surgeon develops world-first program to determine liver transplant success likelihood before surgery

Source: Xinhua 2017-01-12 08:33:38
[Editor: huaxia]

SYDNEY, Jan. 12 (Xinhua) -- An Australian surgeon has developed a world-first computer program to determine the likelihood of success of a liver transplant before the surgery takes place.

The program, developed by a junior surgeon at Melbourne's Austin Hospital, is inspired by the algorithm used by dating websites to pair two people together.

The algorithm uses the top 15 factors known at th time of the transplant to predict the early success of a liver donation and has proved significantly more successful than the current method which is the surgeon's intuition.

The world-first technology successfully predicted graft failure a month after surgery 84 percent of the time, compared to 68 percent for surgeons and researchers are optimistic that it could be adapted for other transplants.

Lawrence Lau, the developer of the program, said he was inspired to develop the machine-learning program by a friend who worked for popular dating website eHarmony.

"When we do the organ retrieval we look at the liver, we feel it with our hands, but ultimately a lot of clinical decisions are based on gut feeling," Lau told News Limited on Thursday.

"At the moment there is not a lot that guides the question of who you give the organ to, other than blood group and the urgency the transplant is required."

Using data from the University of Melbourne, the team from the Austin taught the program to learn using liver transplant outcomes from 2010 to 2013.

Lau said that he hoped the technique could offer thousands of Australians in need of a transplant a second chance at life with fewer risks.

"On one hand if we transplant an organ that is unsuitable, that recipient may die. But sadly, 10 percent of Australians waiting for a liver transplant die while waiting," he said.

"If we could quantify this risk, then we can start maximising the use of this precious resource."

[Editor: huaxia]
010020070750000000000000011100001359755241