Concurrent Multimodal Data Acquisition During Brain Scanning is within Reach

Abstract

Background: Previous brain-scanning research exploring the neural mechanisms underpinning visuomotor planning and control has mostly been done without simultaneous motion-tracking and eye-tracking. Employing concurrent methodologies would enhance understanding of the brain mechanisms underlying visuomotor integration of cognitive, visual, ocular, and motor aspects of reaching and grasping behaviours. Therefore, this work presents the methods and validation for a high-speed, multimodal and synchronized system to holistically examine neural processes that are involved in visually-guided movement. Methods: The multimodal methods included high speed 3D motion tracking (Qualisys), 2D eye-tracking (SR Research), and magnetoencephalography (MEG; Elekta) that were synchronized to millisecond precision. Previous MRIs were taken to provide improved spatial localization. The methods section describes the system layout and acquisition parameters to achieve multimodal synchronization. Pilot results presented here are preliminary data from a larger study including 29 participants. Using a pincer grip, five people (3 male, 2 female, ages 30-32) reached for and grasped a translucent dowel 50 times, after it was pseudorandomly illuminated. The object illumination was the Go cue. Seven discrete time points (events) throughout the task were chosen for investigation of simultaneous brain, hand and eye activity associated with specific visual (Go cue), oculomotor (1st saccade after Go), motor (Reaction Time; RT, Maximum Velocity: MV, Maximum Grip Width; MGW) or cognitive (Ready, End) mechanisms. Time-frequency analyses were performed on the MEG data sourced from the left precentral gyrus to explore task-related changes time-locked to these chosen events. Pilot results: Basic kinematic parameters including RT, MV, MGW, Movement Time, and Total Time were similar to previous, seminal research by Castiello, Paulignan and Jeannerod, (1991), using a similar task. Although no gaze instructions were given, eye-tracking results indicated volunteers mostly gazed at or near the target object when Ready (72%), and then hardly looked away throughout the rest of the task at the important events sampled here (92% - 98%). At the End event, when lifting the dowel, on average, participants gazed at or near the target object 100% of the time. Although saccades > 100 ms after Go, but prior to RT were made on average in about one fourth (M = 13, SD = 6) of trials, a mixed model (REML) indicated their latency in timing after the Go was significantly (F = 13.376, p = .001) associated with RT scores on those trials (AIC = 724, R m2 = 0.407, Rc2= 0.420). Neural activity relative to baseline in the beta band was desynchronized for the visually guided reach periods, beginning prior to Go, and remaining sustained until beyond End, after the grasp and lift were executed. Conclusion: This study presents the layout, acquisition parameters and validation for a multimodal, synchronized system designed to record data from the hand, eye and brain simultaneously, with millisecond precision during an ecologically-valid prehension task with physical, 3D objects. The pilot results align with previous research made with single or bimodal data recordings. This multimodal method enables full-brain modelling that can holistically map the precise location and timing of neural activity involved in the visual, oculomotor, motor and cognitive aspects of reach-to-grasp planning and control.

Kinematic Studies of the Go/No-Go Task as a Dynamic Sensorimotor Inhibition Task for Assessment of Motor and Executive Function in Stroke Patients: An Exploratory Study in a Neurotypical Sample

“…A basic RtG visuomotor task [ 54 ] based on a seminal study by Castiello, Paulignan [ 47 ] was modified to include an increased attentional demand requiring participants to inhibit some reaches. Participants were seated at a desk in a completely darkened room, consistent with the previous study design [ 47 ], with a VPixx DATAPixx [ 55 ] button box located at a midline position.…”

“…Qualisys Track Manager (QTM; Version 2018.1, build 4300) software was utilized to create a model of the hand and dowels, allowing automatic labelling of each marker and manually checking for each participant and task. Once movement data were extracted from QTM [ 56 ], the timeline of task events was imported from the VPixx data files and aligned with the QTM data before importing into a custom processing pipeline created in LabVIEW [ 58 ] to filter the noise (2nd order, low pass Butterworth filter with a cut-off of 0.125 Hz) while extracting key kinematic events [ 54 ]. This processing automatically identified the first marker to move more than 5 mm in each trial in response to the object light on and recorded this as the RT (choosing from either the VPixx button release time, index finger, back of hand or medial wrist), with a minimum threshold of 40 frames (or 160 ms) to ensure a conscious response was captured (as opposed to non-physiological noise).…”