Microsoft wants to use AI to ‘solve’ cancer


Microsoft Corp. has announced several new healthcare initiatives that will put the immense power of artificial intelligence and deep learning into the hands of doctors and medical researchers. The goal? To “solve” cancer.

In a blog post, Microsoft revealed that it has four separate research teams who are using a variety of AI tools to tackle cancer from different angles.

For example, one team is working on generating personalized treatment plans for patients by using natural language processing (NLP) to sift through the wealth of data available in their medical history. Another team is combining machine learning and computer vision to track the progress of a patient’s tumor, rather than relying on the doctor’s eyes alone. A third team is working to create algorithms to better understand how different cancers develop and progress, and the final team is working on a “moonshot” project with the end goal of allowing doctors to program cells to fight a variety of illnesses, including cancer.

According to Jeannette M. Wing, head of Microsoft’s basic research labs, the company’s health project marries computer science with biology to improve the capabilities of both.

“The collaboration between biologists and computer scientists is actually key to making this work,” Wing said.

By the same token, Wing added that the company’s research in programming cells is not only good for cancer research, but also for researching possibilities for biology-based computers.

“If the computers of the future are not going to be made just in silicon but might be made in living matter, it behooves us to make sure we understand what it means to program on those computers,” Wing said.

Microsoft is far from the first company to put AI to work in medical research. A number of doctors have previously turned to IBM Watson for research and medical diagnosis, including a team of Japanese doctors who used Watson to refine their diagnosis of leukemia in a woman whose treatments had so far been unsuccessful.

Innovation vs Profit

In the last few years, we have seen incredible advancements in artificial intelligence research thanks to things like machine learning and Big Data, but developing these sorts of powerful tools is only one step in solving the major problems facing humanity. In a recent interview with theCUBE, Jason Pontin, CEO, editor in chief and publisher at MIT Technology Review, noted that in recent years, many major tech companies have been more shortsighted about their research goals.

“There has been a desire to have a much stronger connection between profit and research,” Pontin said, referring to the trend for major companies to focus on innovations that directly impact their bottom line. Pontin admitted that there are several “oases” of corporate research, such as Baidu and Google’s work on artificial intelligence, but he explained that difficult, long-term research requires a deeper commitment to progress, much like NASA’s Apollo missions in the 1960s and 1970s.

“You can never predict what the benefits of these big projects are going to be,” Pontin said, “but every single time in history we’ve made these big bets, it’s paid off in technologies that ordinary people have used as well.”

Photo by USDAgov