AUTHOR=Gäumann Szabina , Aksöz Efe Anil , Behrendt Frank , Wandel Jasmin , Cappelletti Letizia , Krug Annika , Mörder Daniel , Bill Annika , Parmar Katrin , Gerth Hans Ulrich , Bonati Leo H. , Schuster-Amft Corina 

TITLE=The challenge of measuring physiological parameters during motor imagery engagement in patients after a stroke

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 17 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2023.1225440

DOI=10.3389/fnins.2023.1225440

ISSN=1662-453X

ABSTRACT=It is suggested that eye movement recordings could be used as an objective evaluation method of motor imagery (MI) engagement. Our investigation aimed to evaluate MI engagement in patients after stroke (PaS) compared to physical execution (PE) of a clinically relevant unilateral upper limb movement task of the patients' affected body side.Methods: Twenty-one PaS fulfilled the MI ability evaluation (Kinaesthetic and Visual Imagery Questionnaire (KVIQ-10), body rotation task (BRT), mental chronometry task (MC)). During the experiment, PaS moved a cup to distinct fields while wearing smart eyeglasses (SE) with electrooculography electrodes integrated into the nose pads and electrodes for conventional electrooculography (EOG). To verify MI engagement, heart rate (HR) and oxygen saturation (SpO2) were recorded, simultaneously to electroencephalography (EEG). Eye movements were recorded during MI, PE and rest in two measurement sessions to compare the SE performance between conditions and SE's psychometric properties.Results: MI and PE correlation of SE signals varied between r = 0.12 and r = 0.76. Validity (crosscorrelation with EOG signals) was calculated for MI (r = 0.53) and PE (r = 0.57). The SE showed moderate test-retest reliability (intraclass correlation coefficient) with r = 0.51 (95% CI 0.26-0.80) for MI and with r = 0.53 (95% CI 0.29 -0.76) for PE. Event-related desynchronization and eventrelated synchronization changes of EEG showed a large variability. HR and SpO2 recording showed similar values during MI and PE. The linear mixed model to examine HR and SpO2 between conditions (MI, PE, rest) revealed a significant difference in HR between rest and MI, and between rest and PE but not for SpO2. A Pearson correlation between MI ability assessments (KVIQ, BRT, MC) and physiological parameters showed no association between MI ability and HR and SpO2.The objective assessment of MI engagement in PaS remains challenging in clinical settings. However, HR was confirmed as a reliable parameter to assess MI engagement in PaS. Eye movements measured with the SE during MI did not resemble those during PE, which is presumably due to the demanding task. A re-evaluation with task adaptation is suggested.