AI system uses home videos to aid in Parkinson’s diagnosis

Technology analyzes home-recorded phone videos of patients walking

Marisa Wexler, MS avatar

by Marisa Wexler, MS |

Share this article:

Share article via email
A woman is shown exercising with a brisk walk.

A new system that uses artificial intelligence (AI) to analyze videos of people walking may help speed the diagnosis of Parkinson’s disease.

Scientists at China Medical University Hospital (CMUH) in Taiwan developed the technology, which can help physicians diagnose Parkinson’s more efficiently and get patients faster access to treatment, according to a hospital press release.

The new tool, which analyzes video that patients can record at home on a phone, is “particularly beneficial for patients in rural areas, reducing the need for hospital visits,” CMUH said.

Parkinson’s disease is a neurological disorder characterized by movement problems including bradykinesia (slowed movement), tremor, rigidity, and balance issues. Many patients experience freezing of gait, the temporary inability to move. Patients often describe freezing of gait as feeling like their feet are glued to the ground.

Recommended Reading
banner image for the column

Running for my life: How I exercise to feel better with Parkinson’s

Wait for Parkinson’s diagnosis can mean delay in treatment

There is no biological test to diagnose Parkinson’s disease. A Parkinson’s diagnosis is made by a clinician who confirms that the patient is experiencing motor symptoms characteristic of the disease and rules out other potential causes for those symptoms. It often takes a long time to confirm a Parkinson’s diagnosis, and a wait for diagnosis typically also means a wait to start Parkinson’s-specific treatments that might help relieve symptoms.

CMUH’s new diagnostic system uses AI to identify patients who are experiencing the freezing of gait typical of Parkinson’s disease. AI uses computer data and mathematical algorithms to identify patterns that can be applied to new data.

The CMUH tool can analyze videos recorded on smartphones. Patients can record themselves at home and upload the videos for analysis. The system is able to return results in 30 minutes, CMUH said.

The university said it’s using the system to help facilitate faster diagnosis for Parkinson’s patients.