Kelvin UPDRS

  • Research type

    Research Study

  • Full title

    An evaluation of Kelvin, an artificial intelligence platform, as an objective assessment of the Movement Disorders Society-Unified Parkinson’s disease rating scale.

  • IRAS ID

    266703

  • Contact name

    Thomas Foltynie

  • Contact email

    T.Foltynie@ucl.ac.uk

  • Sponsor organisation

    Joint Research Office

  • Duration of Study in the UK

    2 years, 0 months, 1 days

  • Research summary

    This study aims to help in the objective assessment of the severity of Parkinson's disease.

    Currently a scale called the Movement Disorders Society Unified Parkinson's disease rating scale (MDS-UPDRS) is adopted globally as the best way to measure Parkinson's disease severity, and therefore assess the rate of change and the impact of medication/interventions. The MDS UPDRS has 4 subsections, of which part 3 involves rating the movement ability of the patient for 33 different items including tremor, muscle rigidity and movement slowness in the different body parts. Each of the 33 items is scored between 0 and 4 points therefore the Part 3 subsection has a total maximum score of 132.

    The weakness of the MDS-UPDRS is that different people may rate the scores higher or lower than each other, known as "inter-rater" variability. There is also the issue that some people may rate the same patients differently at different times, depending on human issues such as how busy they might be i.e. "intra-rater variability". This can make it difficult to know whether any change in MDS UPDRS part 3 scores over time is due to an intervention, or simply because there is variability between the people rating the patients.

    To address this, we are using computer technology to create a computer rated MDS UPDRS part 3 score, which we will call "KELVIN UPDRS", based on artificial intelligence/ machine learning from a large number of archived videos of patients.

    For this project we want to prospectively compare the variability in the Kelvin UPDRS with the variability in the traditional human captured MDS UPDRS part 3. It is hoped that the computer derived score will be much more consistent and objective than the human score.

  • REC name

    Yorkshire & The Humber - Sheffield Research Ethics Committee

  • REC reference

    19/YH/0421

  • Date of REC Opinion

    3 Dec 2019

  • REC opinion

    Favourable Opinion