IBM—after February’s Jeopardy debut of its deep question-answering technology, Watson—announced partnerships with medical software application providers, medical professionals and hospitals to create applications using the technology to advance medical practices. But that’s not all it’s lined up to do.

Eric Nyberg, professor of the language technologies institute at the Carnegie Mellon school of computer science, led the CMU team of students and faculty in the development of the Open Advancement of Question Answering Initiative (OAQA) methodology with IBM. CMU made two direct algorithm contributions: a source expansion algorithm and an answer-scoring algorithm. Both of these technologies promote Watson’s use in a number of industries, mainly because of its ability to be a self-contained database.

“Any area where you have folks who have to sift through lots of textual material to either support or refute a hypothesis, or to even formulate a hypothesis; those will be the areas where Watson will make a difference,” Nyberg said in an interview before a symposium held at the University of Pittsburgh in March.

IBM, along with doctors at Columbia University Medical Center and the University of Maryland School of Medicine, will apply Watson’s technology to create a sort of “doctor’s assistant,” according to IBM.

In February, IBM also announced that they will team up with Nuance Communications to create medical software programs that can use Watson to store a variety of medical journals, research papers and other relevant data without calling to the Web for every search transaction.

Nico Schlaefer, a Ph.D. candidate at CMU who worked on Nyberg’s team, said that Watson’s technology could take on a natural-English question (such as “Where can I travel this weekend?”), and it will return relevant rates based on that question instead of just basing answers on keywords (like “travel” or “weekend”). This, he explained, is what separates it from Google’s search.

At CMU’s symposium, both Nyberg and Don Burke, University of Pittsburgh’s dean of the graduate school of public health, said professors could also use the technology.

For law professors, as an example, law journals, research papers and past decisions could be stored in a repository powered by Watson, and the algorithms contributed by CMU and IBM could then sift through the data to return requested information almost immediately, as Watson did in the Jeopardy contest.

Jim Spohrer, director of university programs for IBM, said government agencies could use Watson’s technology to organize assistance programs as well as tax information programs.

He said that government agencies, such as those that regulate healthcare, retirement benefits or welfare benefits, could store their information in Watson for access by the public. Because of Watson’s ability to understand natural language questions, it could be used to replace or update customer service phone systems or Web-based inquiry sites.

Data organization
Burke, in his interview with IBM at the symposium, said that the problem many physicians face today is in integrating the data they have and determining a diagnosis. There are so many things that factor into the health profile of a patient, including genetics, current health and health history, which then need to be matched with possible diagnoses. The more organized this data is, the quicker results can be determined by medical professionals.

He added that the next step for technology in the public health sector could be with data organization for countries as well.

“Most countries around the world still have information about their diseases that need work,” Burke said. Not enough of their data, he added, is organized or has been collected yet. He said that Watson’s technology could be used to organize known diseases, symptoms and other essential information.