Purdue researchers introduce novel machine-learning technique to detect age-related changes in tremors

Headshots of Anne Sereno and Aditya Shanghavi

Anne Sereno (left) and Aditya Shanghavi (right)

Typical human aging brings about subtle changes in the body that can impair daily activities and quality of life. For many people, aging brings about tremors and the slowing of hand movement.

A pioneering study from Anne Sereno, professor of psychological sciences in the Purdue University College of Health and Human Sciences and biomedical engineering in the Weldon School of Biomedical Engineering, and her laboratory has paved the way to accurately identifying subtle age-related tremors. The study titled “A Machine-Learning Method Isolating Changes in Wrist Kinematics that Identify Age-Related Changes in Arm Movement,” was published in Nature: Scientific Reports.

Using wearable sensor technology and innovative machine-learning techniques, the research team, led by Aditya Shanghavi, a Biomedical Engineering PhD candidate, analyzed the wrist kinematics of younger and older adults performing standard clinic-based tasks and identified kinematic variables that accurately and reliably distinguished healthy older adults from their younger counterparts. Accurately identifying normal age-related tremors is critical so they don’t interfere with the diagnosis of tremor disorders in older adults.

“Tremors can be exacerbated by food, medications, and even sleep, making the development of objective, repeatable and portable measures of tremors key for more reliable assessments,” Shanghavi noted.

The sensitivity and accuracy demonstrated by the novel data-driven methodology paves the way for a range of applications, including isolating changes in motion across various body parts and conditions and facilitating early detection of tremors in neurological diseases, such as Parkinson’s disease.

“These findings suggest many possible exciting future directions, such as enhancing current subjective evaluation approaches in the clinic or making possible telehealth and treatment monitoring outside the clinic,” Sereno said.

Source: Weldon School of Biomedical Engineering