07 October 2016

Computer for early diagnosis of parkinsonism

Working on the keyboard will help identify Parkinson's disease

Oleg Lischuk, N+1

American and Spanish scientists have developed and tested software for early detection of Parkinson's disease by typing on the keyboard. The results of the work are published in the journal Scientific Reports (Giancardo et al., Computer keyboard interaction as an indicator of early Parkinson's disease).

Parkinson's disease is the second most common neurodegenerative disease (after Alzheimer's disease), which affects up to 10 million inhabitants of the Earth. It is manifested by tremor, stiffness of movements and other disorders. Their development is associated with the death of dopamine-producing neurons in the motor centers of the brain, primarily the striatum and the tire. It is currently impossible to cure Parkinson's disease, but timely therapy can slow down the progression of symptoms for a long time. Therefore, early diagnosis plays a key role in the success of treatment. However, existing methods do not allow monitoring the development of the disease with sufficient frequency and coverage.

To remedy this situation, the staff of the Massachusetts Institute of Technology has created a program that tracks the time of pressing the keys (the interval between pressing and releasing, usually about 100 milliseconds) when typing. They decided to use it to detect Parkinson's disease based on the fact that neurologists for the same purpose use tests of pressing one key and tapping with thumb and index fingers, technically similar to working on a keyboard.

The Americans tested the effectiveness of their development in two Madrid clinics. The study involved 42 patients with early-stage Parkinson's disease and 43 healthy volunteers included in the control group. All participants were asked to type random text for 10-15 minutes on computers with an experimental program installed.

It turned out that in patients with Parkinson's disease, the time of pressing the keys varies greatly, while in healthy volunteers it is more or less constant. The analysis of these data by a specially created algorithm allowed to identify a certain "handwriting" characteristic of early Parkinsonism.

The use of this "handwriting" made it possible to distinguish people with the disease from healthy people with high accuracy. The area under the ROC curve (ROC AUC is an indicator that quantifies the suitability of the technique for distinguishing carriers and non–carriers of any trait) for this method was 0.81. Thus, its diagnostic value turned out to be higher than that of the tests of pressing one key (ROC AUC 0.61) and tapping with thumb and index finger (ROC AUC 0.75).

The developed diagnostic program can be installed on any computer, added to the firmware of the device or placed on a web page to monitor and quantify possible symptoms during daily use of the computer. To ensure confidentiality, the typed text is not analyzed and is not saved by the program.

This is not the first time that computer work has been proposed to be used for the diagnosis of neurodegenerative diseases. So, in March 2016, scientists reported that rare computer use can serve as an early sign of the development of dementia, for example, in Alzheimer's disease.

Portal "Eternal Youth" http://vechnayamolodost.ru  07.10.2016

Found a typo? Select it and press ctrl + enter Print version