Smartphones to detect early stages of Parkinson’s disease

A novel new way to diagnose Parkinson's disease may involve nothing more complicated than carrying a smartphone in your pocket.

Researchers at Aston University have developed a phone app that can spot signs of the disease with more than 90 per cent accuracy.

The phone provides a range of checks including a voice analysis, monitoring of the person's gait and also dexterity tests using features already provided on most smart phones, said Dr Max Little of Aston's Nonlinearity and Complexity Research Group.

Early diagnosis is very difficult, he said. “They don’t know how to detect the disease before it is too late when symptoms arise.”

His app is grounded in work he began back in 2006, when given a block of voice recordings provided by Intel.

Minor “voice impairments” represented a possible means of early diagnosis and by finding a way to detect these he managed to reach 86 per cent accuracy in diagnosis.

Things have progressed much further however with the appearance of smart phones.

These can provide high definition voice recordings for analysis but they also feature built-in “accelerometers”, used to keep the the screen display upright.

However, when the phone is used as a diagnostic tool, the person is asked to pronounce a sound for the phone recorder, walk 20 paces so the accelerometers can record movements and then the person completes simple finger dexterity tests on the touch screen.

Results so far are “promising”, he said. “The tool is extremely accurate.”

The phone app is able to give objective data for doctors that can be used to support decision making on treatment options in the clinic, Dr Little said.

He is involved with Oxford University on a joint study of 2,500 subjects and five years into the project he recently received funding for a further five years.