Phone: (416) 785-2500 ext. 2017
E-mail: sgraham@rotman-baycrest.on.ca
The Rotman Research Institute,
3560 Bathurst Street, Room 1060
Toronto, Ontario, M6A 2E1
Phone: (416) 785-2500 ext. 2080
E-mail: volunteers@rotman-baycrest.on.ca
Brain activity generated in VR must be well characterized to develop optimized VR physical therapy for stroke patients. Typically, those that advocate performing behavioural tasks in VR implicitly assume that the brain activity under these conditions is very similar to that elicited in the real world. By extension, tasks performed in VR, and improved task performance in particular, are anticipated to generalize to those in the real world. However, there is a great variability in the "level of realism" achievable in virtual environments, depending on the sensory stimuli available. As physical therapy for stroke patients often involves regaining control of the hand, VR / fMRI experiments are to be conducted that look at the effect of increasing the realism in VR by successively adding multimodal sensory input (visual and somatosensory stimuli) to a data glove device, and then comparing the brain activity generated when using this device to that obtained for hand-object interactions in the real world. The interactions are very simple and reductionist, involving tapping or moving virtual objects. This work should shed light on how well VR can be used to elicit brain activity that is similar to that obtained associated with real-world actions. Experiments are to be conducted on young and elderly healthy adults, as well as patients in the chronic phase of stroke recovery. As a consequence, we will also be able to characterize the effect of normal aging on brain function in VR, what level of realism in VR is well tolerated by stroke patients, and whether VR has applicability in a functional neuroimaging context to understand better the consequences of brain damage due to stroke.
1) Hand-object interactions in VR (touching and manipulating virtual objects) will exhibit brain activity patterns similar to those associated with very similar actions in the real world, only if appropriate multimodal visual and somatosensory stimuli are provided in combination: visual display of hand/finger/object interactions correctly in three dimensional space; correct proprioceptive stimuli; and force feedback to provide somatosensation of virtual surfaces.
2) Given the sensory loss that accompanies normal aging, normal elderly subjects will have impaired ability to perform hand-object interactions compared with normal young adults, and will exhibit increases in brain activity consistent with the need to recruit more neural resources to complete the tasks. As a consequence, elevated brain activity, and possibly the recruitment of additional brain regions, will be observed across tasks in VR and in the real world in comparison to the activity observed for young healthy adults.
3) Patients in the chronic phase of stroke recovery will be able to tolerate VR exposure, depending on their specific deficits. Those able to perform hand-object interactions well will show patterns of brain activity consistent with what is known about the brain's ability to reorganize in response to focal damage. Those performing the tasks poorly will reflect brain activity consistent with the size and location of their lesions.
Structural MRI refers to a collection of techniques that visualize brain anatomy, whereas fMRI is a newer method that provides pictures of brain function. Functional MRI visualizes the activity of regions of neurons in terms of the localized changes in blood flow, blood volume, and blood oxygenation that occur when the brain performs specific tasks. Functional MRI has tremendous potential as a scientific research tool to increase our understanding of how the brain works.
My area of research is medical imaging physics and its applications. During my Ph.D. at Sunnybrook Health Sciences Centre, I studied the fractal properties of vascular structures and models of organ blood flow using computed tomography. In that time, I became more and more interested in human cognition and the brain. Consequently, after completing my Ph.D. research, I joined the Rotman Research Institute as a postdoctoral fellow in December 2006 to work on the advancement of brain imaging techniques.
Two research projects are currently in progress:
Both fMRI and MEG provide measures of brain activity. However, fMRI signals are related to changes in local blood flow while MEG records changes in the magnetic field around the head related to changes in neuronal currents in the brain. In addition, a feature of fMRI is high spatial resolution with poor time resolution while MEG, in comparison, has lower spatial resolution and higher temporal resolution. Combining these two techniques offers therefore a more complete picture of brain activity than either technique by itself.
In the current project, we are using both of these techniques to measure the response of the brain to vibrational stimuli on a finger.�The stimuli cause a neuronal response in the brain, associated with changes in the MEG signal, which then causes an increase in blood flow and a subsequent change in the fMRI signal. Based on our knowledge of both fMRI and MEG signals we are trying to improve our current understanding of this chain of events. This will improve our ability to detect changes in normal brain function, for example due to stroke. In stroke patients, the sense of touch is often impaired. Such patients will therefore be part of this study to investigate how the response of the brain to a vibrational stimulus has changed. This, in turn, may provide clues for future therapies.
In recent years, a new technique in MR imaging called Parallel Imaging has created the potential to acquire images much faster than previously possible. However, this is also associated with a loss of image quality. In this study, we are comparing different levels of acceleration based on commonly performed measures of brain morphology, for example measures of local brain volume. If the change in these measures is minor, then acceleration is recommended. However, major changes would discourage acceleration.
Acceleration is possible for a number of sequences that are commonly employed clinically. The reduction in scan time benefits the comfort of patients and reduces the cost for their examination.
Developing Multi-Echo fMRI for Neurofeedback Applied to Motor Learning
My research involves the development of an investigative and potentially therapeutic tool involving functional magnetic resonance imaging (fMRI), and will lead to the creation of an imaging framework and experimental infrastructure for future application in stroke rehabilitation research.
Stroke is one of the leading causes of mortality and disability in Canada, and impairment can lead to a range of cognitive and physical disabilities. My work targets motor learning, with the long-term goal of facilitating motor recovery and improving the quality of life after stroke. With neuroimaging, developing therapeutic techniques can incorporate neurological assessments of recovery and learning, including a method that attempts to enhance learning behaviour through direct modulation of brain activity.
I put forth a general hypothesis that neuroimaging-mediated biofeedback training can produce positive behavioural changes in healthy adults. It includes the technical development and basic neuroscience investigations necessary to characterize the method, and establish its efficacy for rehabilitation. It is the first step towards the exploration of its use in motor therapy and for further examination of the mechanisms of learning and motor recovery after stroke. This research spans a broad spectrum of work in fields of electroencephalography (EEG) biofeedback and early functional magnetic resonance imaging (fMRI) biofeedback, as well as research in mental chronometry and the study of motor networks.
In working to address the stated goals, I am currently involved in many technical aspects of fMRI acquisition, including:
Infrared Position Tracking for fMRI Motion Correction and Optimization
The magnetic resonance imaging (MRI) modality is very sensitive to motion artifact. Particularly in functional MRI where subjects are assessed according to cognitive tasks and stimuli and are more likely to move. Currently, there are many passive methods to correct for motion using post processing algorithms. My research is to optimize post processing motion correction using co-registered position tracking data. This will be a more active approach to correcting for motion that is measured in real-time while the MRI machine is scanning.
The research revolves around infrared sensitive cameras that can be used to track markers in 3D space. Just like our eyes, multiple cameras (verses a single camera) imaging the same field of markers can be used to obtain position data in three planes using triangulation for depth. Infrared sensitive cameras are used so that we can control what objects reflect in that spectrum and isolate highly reflective markers. We also have the added advantage of imaging in low light.
In addition to position tracking solely for the purpose of MR image correction (head movement, we can also look into collecting kinematic data for use in conjunction with fMRI studies. My research will also focus on the development of different camera setups and tracking tool designs to capture movement of other parts of the body. Certain tasks like pinching fingers and moving toes can be detected with the infrared cameras and co-registered with the fMRI images of the brain. This is intended to give neuroscientists more versatility and accuracy with their data analysis.
Challenges:
Grigori Yourganov, PhD. Student
Nathan Churchill, M.Sc. Student
Surgical Spotlight ~ Dr. Tom Schweizer
Parisa Bahrami, M.Sc. Student
Karen Kan, M.Sc. Student
Baycrest is an academic health sciences centre fully affiliated with the University of Toronto
Privacy Statement | Disclaimer | © 2012 Baycrest. All Rights Reserved.