Researchers at Indiana University-Purdue University Indianapolis recently developed new computer technology that can help patients suffering from Acute Myelogenous Leukemia, or AML.
Oncologists estimate about 20,000 people in the U.S. were diagnosed with this rare blood cancer last year. But it could become more common as the population ages.
The situation of AML patients is not hopeless: The cancer is curable, but treatment can be long and painful and patients can relapse.
"It's like two colleagues working on the same problem. One is a human being, the other is a computer."
IUPUI Associate Professor of Computer Science Murat Dundar says the computer model provides another tool for determining whether treatment will be effective.
"It's like two colleagues working on the same problem. One is a human being, the other is a computer," Dundar says. "So the physician does his own work and the computer works in the background. Once the computer processing is over the physician can invoke those results from the computer and can double check whether he agrees."
That's how Dundar and his colleagues envision their computer aiding system in a clinical setting. The scientists fed the computer with bone marrow data and medical histories of AML patients, as well as blood data from healthy individuals.
"It turned out that the computer actually did a really good job," Dundar says. "It predicted remission with 100 percent and relapse with 90 percent accuracy."
Even though the algorithm turned out to be quite successful, the researchers say physicians can't be replaced by a computer.
"It's like a second reader working on the same case and trying to make sense of data," Dundar says. "This way we want to reduce mistakes and misdiagnosis to the extent possible."
So far the model is just a concept study; implementing it into the daily routine of clinics will not start for the next three to five years, depending on how quickly the scientists can get funding.
