The findings, published in the journal Science Translational Medicine, suggest the platform could allow clinicians to remotely monitor the progression of a patient's condition and adjust medication plans accordingly to improve outcomes.

Parkinson's disease is marked by a breakdown in voluntary movement (dyskinesia) and the appearance of tremors, which severely detract from a patient's quality of life.

"These symptoms can be treated with medications, but patients respond best if clinicians can precisely titrate and change drug regimens to match the severity of symptoms," said the researchers, including one of Indian-origin Adeeti V. Ullal, Senior Manager, Motion Health Technologies, Apple.

"However, clinicians currently base their assessments of patients on infrequent clinical visits that fail to capture subtle changes in symptoms," they added.

The team tackled this issue with their Motor fluctuations Monitor for Parkinson's Disease (MM4PD) -- a collection of algorithms that uses smartwatch sensors to capture daily fluctuations in the wearer's movement patterns.

The system identified patterns in the severity of tremors and dyskinesia over six months when worn by 225 patients and detected changes in symptoms that clinicians noted might be missed in traditional evaluations.

Furthermore, MM4PD recorded signs of emerging tremors and impairments, which may require changes in medication schedules to treat properly, the team said.

For the study, the system was tested in a study involving 343 patients -- including 225 who the researchers followed for six months.

The system gave evaluations that matched a clinician's estimates in 94 per cent of the subjects.