Park Test

2021: Service Launch

At the end of November 2021, information appeared that the software could predict with great accuracy the probability of developing Parkinson's disease in a person taking a photo or selfie with the same reliability as wearable digital biomarkers that track motor symptoms. The Park Test service is based on computer vision technology and includes various types of algorithms.

A reaction in the form of a smile on a person's face, not the only behavior that the software can analyze for early symptoms of Parkinson's disease or related disorders. The researchers developed a five-component test that neurologists can perform with patients sitting in front of their computers' webcams hundreds of kilometers away. Scientists described the essence of the technology in the journal Nature Digital Medicine, the publication is the most cited research journal in the field of preclinical medicine.

Park Test service launched, identifying signs of Parkinson's disease by selfie

According to Esan Hoke, associate professor in the Department of Informatics at the University of Rochester, this can make a huge difference for patients who are quarantined, immobile or living in underdeveloped states where access to a neuropathologist is limited. In addition to the fact that patients taking the test should not only smile and alternate the smile three times with a neutral facial expression, they are also asked to:

• read aloud the complex written proposal;
• 10 times quickly touch the forefinger to the thumb;
• to make the most disgusting expression of the face, alternating it with a neutral expression, three times;
• raise the eyebrows as high as possible, then lower them as far as possible, three times slowly.

Using machine learning algorithms, a computer program for several minutes shows the percentage probability that according to the results of each of the tests, the patient shows symptoms of Parkinson's disease or related disorders. When a patient smiles, the PO can determine whether their control of the facial muscles is reduced at the same time, this is one of the symptoms of Parkinson's disease, which doctors call modularity. One of the features of the disease is that not all symptoms appear constantly, and not all symptoms appear in each part of the body, hence the importance of testing other expressions and movements.

Researchers have made great progress in identifying Parkinson's disease by automatically analyzing facial expression, voice, and motor movements. However, further work is needed to develop algorithms to distinguish involuntary tremor from other motor disorders, including ataxia and Huntington's disease. A team of scientists aims to differentiate these tremors with artificial intelligence (AI) to prevent potential harm from misdiagnosis while maximizing benefits. Although ethical and technological issues remain to be resolved, the Gordon and Bettie Moore Foundation has agreed to fund the study. Collectively, the researchers' efforts contribute to a future in which equality and access to neurological care will be as ubiquitous as owning a smartphone or other device with Internet access .^[1][2]

Notes