Can action observation and/or motor imagery be used to improve computer-based actions in people with Parkinson’s?

Introduction: Parkinson’s is a neurodegenerative condition characterised by motor symptoms such as resting tremor, bradykinesia and rigidity. Many people with Parkinson’s (PwP) experience difficulty when using a computer (e.g. Cunningham et al., 2012) such as reduced keyboard/mouse speed, difficulty clicking or controlling a mouse and hitting accidental keys when typing. Further, PwP are interested to improve computer-based actions (PUK survey). Computer use is becoming an ever more central part of daily life, particularly due to the 2020 pandemic; hence it is crucial to determine if low cost tools could be developed to facilitate computer use. Growing research has indicated that watching others’ actions (action observation; AO) and imagining actions (motor imagery; MI) can be effective tools for facilitating movement in PwP (e.g. Bek et al., 2018). Further, recent studies have suggested that combining the two techniques (AO+MI) could be more powerful in improving movement execution than either one alone (Bek et al., 2019). Therefore, the aim of this study is to determine whether AO+MI can facilitate computer-based actions in PwP, and whether this is more effective than MI alone. 
Method: An online within-subjects study design will be utilised with PwP completing the experiment from home. The experiment will consist of two sections: 1) typing, and 2) computer mouse use. In 1) PwP will be asked to type a word, and in 2) they must click a target with an on-screen cursor. Using tailored measurements, movement speed, accuracy and motor control will be examined across three different conditions: a) AO+MI, b) MI alone, and c) control. The UPDRS, KVIQ and coin-rotation tasks (CRT) will also be administered as measures of Parkinson’s symptoms, motor imagery ability and dexterity respectively. 
Approach for statistical analysis: Multivariate linear mixed models will be used to analyse measures of speed, accuracy and fine motor control across conditions a), b) and c), in both typing (1) and computer mouse tasks (2). Regression analyses will be conducted to investigate whether task improvements on 1) across conditions a), b) and c) predict improvements on 2), as well as whether scores on the UPDRS, KVIQ and CRT predict performance differences across conditions