By Rob Knies
February 6, 2010 11:00 AM PT
Eric Chang has a background in speech recognition, while Shipeng Li’s work focuses on multimedia analysis. Yet neither Microsoft Research Asia computer scientist is surprised to find himself working on ways to improve healthcare for people worldwide.
That work was the focus of the Microsoft Research Asia eHealth Workshop 2010, held Feb. 4-5 at the Beijing Sigma Center. About 150 people attended the event, which gave researchers from a variety of disciplines an opportunity to share the results of their work and to meet others with whom new, valuable collaborations could be launched.
The workshop gave Microsoft Research personnel a chance to discuss related research and activities—and for invited members of university faculties to offer insights into their efforts, all with the intention of addressing computing trends that are expected to transform the delivery of healthcare services in the coming decade.
The potential for collaboration, said Chang, director of Technology Strategy for Microsoft Research Asia, was the driver behind the event.
“The goal,” Chang explained, “was to bring together people from academia and industry, in multiple different research areas related to healthcare, to foster discussion, to learn from what each is doing, and to create new collaboration opportunities.
“We believe that technology will play an increasingly important role in getting people healthier.”
And, he added, such a technological approach can originate from surprising directions.
“Previously, I worked on speech recognition,” Chang said. “In fact, one of the earliest uses for speech recognition was to help doctors who dictate medical reports, such as radiologists who enter their medical reports from reading X-rays.
“There’s a lot of technology for extracting knowledge from data—for example, natural language processing and machine learning—that can be applied to healthcare. Looking at medical data, there is a lot similarity between those domains and healthcare.”
Li, principal researcher and manager of Microsoft Research Asia’s Media Computing Group, uses the example of Chinese medicine to connect his multimedia-analysis work with efforts to improve healthcare.
“In traditional Chinese medicine, an experienced doctor can tell by touch what’s going on in your body, whether something is wrong,” he said. “The body is viewed as a system. I’ve been thinking, ‘What if we had the medical data for many, many people? We could infer from their symptoms what’s going on in a person’s body.”
Such computing-based interests are buttressed by natural, human concerns.
“It’s also, in part, relative to age,” Chang explained. “As you get older, your loved ones go to the hospital more often, or you yourself go to the hospital more often. You start to see the importance of keeping healthy.
“Healthcare is what I’d call a ‘never-complete’ problem. Right now, in the U.S., it’s not unexpected for someone aged 75 to have a hip replacement in order to continue to jog or play tennis. Years ago, that probably would not even have been considered. As our treatment technology improves, people’s expectations improve, as well. Imagine that, someday, our technology is good enough for people to live to 120, but a new treatment or method comes along that enables a lot of people to live to 150. Our expectations will also rise. People’s expectations are never satisfied.”
Of course, the possibilities engendered by the eHealth Workshop dovetail with the vision of Microsoft, including the company’s Health Solutions Group: to improve health around the world through software innovation.
In particular, the workshop focused on the approaching revolution in healthcare services stemming from recent trends such as the need for continuous increases in effectiveness and efficiency; the utilization of technology to collect data that can improve health, especially for aging societies; and harnessing computing’s analytic abilities to assist physicians.
“We believe,” Chang said, “a multidisciplinary, joint effort will be required to enable a huge amount of progress.”
Thus, the workshop, which also featured a session of project posters, addressed a broad swath of research areas, including data mining, mobile devices and sensors, natural language processing, signal processing, data visualization, and medical imaging.
Li provided an example of how his research focus could be applied to healthcare advances.
“We are doing a lot of multimedia content analysis,” he stated. “We take multimedia data, analyze its low-level features, and abstract with some human annotation to identify semantic, high-level features that could be useful for getting more information out of raw data.
“Healthcare is very similar. We use different sensors to collect data from the human body, and, sometimes, we use environmental sensors or environmental data. We put all this together. The low-level data itself might not mean much, but if you add data analysis similar to that we apply to multimedia data, we can abstract meaningful, high-level semantics that can be used to help patients understand what’s going on with their bodies or to help doctors take a look at a user’s history.”
The workshop, organized by Miran Lee, senior manager for the Microsoft Research Asia University Relations team, featured a collection of leaders in their respective fields who, after opening remarks from Hsiao-Wuen Hon, managing director of Microsoft Research Asia, heard Chang, Li, and Wenwu Zhu, senior researcher, discuss pertinent projects from their lab. Lolan Song, senior director of University Relations, outlined her team’s efforts in academic collaboration. And a series of 12 talks demonstrated the breadth of the event’s approach:
In addition, right before closing remarks from Chang, the workshop featured a panel discussion on Challenges and Opportunities in eHealth, moderated by Chang and featuring:
The organizers of the workshop hope the workshop provides benefits to both Microsoft Research Asia and the greater healthcare research community.
“There are a lot of opportunities for people here at the lab,” Chang said, “to work more in healthcare research if they become more aware of the type of problems being faced and the opportunities that exist in this field.
“Ultimately, though, we hope to foster interesting collaborations. If a professor in Japan and a professor in Korea decide there’s something they can work together on, that would be a nice outcome.”
What all involved hope to effect, however, is an improvement in the health of people the world over—sooner rather than later.
“Our goals are to save lives,” Chang concluded. “Essentially, how we measure ourselves will be whether we have impact into creating systems and methodologies that will improve people’s lives in a five-year, 10-year time frame.”