September 29, 2011
“The big dream,” says Konrad Kording, “is to detect Parkinson’s before it even happens.” The 38-year-old sips herbal tea as he talks, his German accent barely noticeable after seven years in America. The skateboard he rode to work is propped outside of his 14th-floor office at the Rehabilitation Institute of Chicago.
At the unassuming office building in Streeterville, Kording—an associate professor at Northwestern University’s Feinberg School of Medicine who earned his Ph.D. in Switzerland—is one of hundreds of inventors working to help patients suffering from physically debilitating ailments. One floor below, a team has been working for years to develop a mind-controlled prosthetic leg.
Kording, along with Dr. Santiago Toledo, is father to one of the institute’s newest projects: a smartphone program designed to help diagnose and monitor Parkinson’s, a degenerative disease of the nerve cells that afflicts more than 1.5 million Americans.
Insert requisite joke here about there really being an app for everything these days. But at its core, this program isn’t as far from Angry Birds as you may think. Shortly after the iPhone was launched, Apple debuted a series of TV ads in which people holding phones swapped contact information with a simple fist bump. The movement sensor that makes this possible, called an accelerometer, is also how a game like iBowl lets you knock over pins with the flick of a wrist. Over coffee last year, Kording and Toledo, medical director at the institute’s Orthopedic Rehabilitation Program, had a thought—what if the same technology that lets you “swing” an onscreen nine-iron could collect and chart the movements of Parkinson’s patients?
Many of us were introduced to Parkinson’s via Michael J. Fox, who was diagnosed in 1991 at age 30. But the disease—characterized by uncontrollable shaking or trouble balancing; basically, your central nervous system is failing—is rare in people younger than 50. The average patient is more like Bucktown resident George Pappageorge, 57, one of about 20 people testing the app Kording and Toledo conceived.
Diagnosed six years ago, Pappageorge speaks slowly and methodically about living with Parkinson’s. He went to a physician after noticing that his left leg “felt funny” and he was having trouble squeezing toothpaste on his toothbrush. The disease, which runs in his family, mostly affects his gait; his left leg tends to drag. It hasn’t prevented him from working as an architect but does make it more difficult for him to play golf.
Pappageorge came to the Rehab Institute this spring after hearing about amplitude training, a style of voice therapy for Parkinson’s, from a woman seated next to him on a flight home from California. At the institute, Toledo invited Pappageorge to try the smartphone app. He decided to help test it because he wants doctors to better understand his disease. “All patients are different and progress at different speeds, and it’s confusing to find out how to administer services and medications,” Pappageorge says.
To train the app to respond to his body, Pappageorge selects an action, such as sitting or walking, and presses start. He slips the phone in his pocket and the app measures his movements until he takes it out and presses stop. Eventually, Pappageorge performs a series of actions without stopping the program, seeing if the app can recognize what he’s doing. In a study with seven patients, the app was able to identify what activity a patient was performing with 95.2 percent accuracy.
Out of the lab, patients simply carry the phones in their pockets; the app, which can stay turned on all day, is constantly collecting data. When patients upload the data, doctors can detect tremors, falls and more. By comparing that information with pill dosages and combinations, the app could help identify whether patients’ medications are working. Toledo, Kording and the app’s builder, NU postdoctoral researcher Mark Albert, believe it will prove much more accurate than having patients journal every hour, a common technique to monitor Parkinson’s.
It could also, Kording says, change the lives of at-risk patients—those with a family history of Parkinson’s or who have suffered head trauma. The team’s goal is to fine-tune the app to pick up on movements so nuanced, it will be able to diagnose Parkinson’s long before a doctor could, giving a leg up on treatment.
Kording and his colleagues are hoping to present their work at conferences this fall and have the app available to doctors and patients within a year. In the meantime, Pappageorge’s voice therapy is paying off: He no longer needs to walk with a cane. As he moves down the lab’s hallway for one of the day’s final exercises, he makes a complete turn and begins walking backward. “I couldn’t do this before!” he hollers, a big grin on his face.