Motor assessment in Parkinson’s disease is plagued by issues of subjectivity and reproducibility – not to mention systematic biases such as rater drift. To counter this, Machine Medicine are developing a platform, called Kelvin, that can be used from any device (e.g. smartphones or tablets).

The platform allows researchers to record, score, store and analyse video of Parkinson’s patients during motor assessment.

Once captured and organised, the platform uses a cutting-edge computer vision technique, called pose estimation, to objectively track the movement of the patients down to the fingertip level and deliver objective metrics of core symptoms, such as tremor and bradykinesia.