Hundreds of new cancer drugs are being developed and new research is published every minute to help physicians to develop specific treatments based on the constituent elements of the illness of each patient.
The problem is that there is too much material to read and many drug combinations so that doctors are able to choose the best option for each case.
In comes a machine learning project from Microsoft Research called Hanover, which aims to digest all scientific papers and help predict which drugs and which combinations are most effective, according to the company.
Researchers at the Cancer Knight Institute at the University of Health and Science Oregon, are working with the architect Hanover, Hoifung Poon, to use the system in order to find effective combinations of drugs to combat acute myeloid leukemia, a type of cancer often deadly whose treatment makes little headway for decades.
Among the researchers are Jeff Tyner and the director of the institute, Brian Druker, best known for having pioneered with Gleevec, a successful drug for another type of leukemia that now belongs to Novartis and helped double the survival rate five years these patients since the 1990s.
Cancer is caused by genetic mutations that cause cells to grow and multiply in an uncontrolled manner.
To better identify these specific mutations enabled the emergence of new drugs to fight the disease more accurately, increasing the chances of survival.
More than 800 drugs and cancer vaccines are in clinical testing phase, according to a report in 2015 the Pharmaceutical Research and Manufacturers of America.
At the same time, increasing the speed and reducing the cost of sequencing the genes increased the amount of research and development treatments, which means that most cancer patients can obtain data on its specific case.
"This is exciting, but it is also challenging to define what to do with all this information," Tyner said, the Knight Institute.
"That's why it is so important the idea that a biologist working with information scientists and computer. The combination of all these features help to make great strides in the development of more efficient and less toxic treatments. "
Hanover is one of several projects announced Tuesday by Microsoft to develop computational approaches to improve treatment and cancer research.
Other projects involve machine learning and computer vision system to help radiologists to understand the evolution of tumors and an initiative that could one day allow scientists to program cells to fight disease.
The Hanover project, Poon intends enhance the work of so-called tumor review boards, in which several doctors meet to discuss the best treatment option for patients.
"One of the current bottlenecks in the tumor advice is to understand all this knowledge and define how to go beyond," he said. "That's what people have to deal with, unless we can automate this process."
Nenhum comentário:
Postar um comentário