BrainsCAN Fellows

Research Information:
The corpus callosum is a large nerve fibre bundle that connects the left and right cerebral hemispheres, and a reduction in its size impairs cognitive tasks that involve integrating information from multiple brain regions. Adults with Autism show a decrease in corpus callosum size when compared to controls. However, the exact mechanics of resulting impairments, at both a structural and functional level, are not yet fully understood. Using tools from graph theory, a branch of mathematics that focuses on formally modelling the structure and features of networks, we can analyze digital reconstructions of both structural and functional neural networks. By quantifying the properties of these networks, we can find measurable differences in the network structure across health and disease, allowing for a better understanding of how differences in brain structure cause differences in neural activity. Through computational analysis and modeling, we will also investigate how network connections can give rise to spatiotemporal patterns of neural activity. This could help explain symptoms of the disorder by shedding light on how and why neural networks in patients with Autism respond differently to stimuli than those of controls.

Research Information:
Every culture has rhythmic music with a beat: the periodic underlying pulse that structures our temporal perception and leads to spontaneous movement. Previous studies show that although the auditory system is predominant in perceiving the temporal sequences, this task can potentially be achieved via other modalities like vision and touch, and there exists a common neural network for beat detection in the three above mentioned modalities. However, with many investigations in timing and beat perception, the question of what helps beat perception has yet to be elucidated. We assume that there is a hierarchical structure between the three levels of timing ability including single duration, nobeat sequences, and beat-inducing sequences so that successful performance in each level provides a chance to proceed to the next step, and the same perceptual hierarchy is present across three sensory modalities of audition, vision, and touch. We will address this gap in the literature using the temporal discrimination task based on the reminder task paradigm. The outcomes will be valuable in some ways. Firstly, it will provide a means of investigating the cognitive mechanism underlying beat perception. Moreover, this study will inform how temporal processing occurs in different modalities.

Research Information:
Parkinson's disease (PD) is the second most common neurodegenerative disorder after Alzheimer's disease, affecting 1 in every 500 Canadians. Individuals diagnosed with PD typically exhibit a distinctive profile of progressive impairments.Generally, PD affects motor functions. However, in many cases, PD patients exhibit a mixture of both motor and cognitive impairments. Our ability to predict these distinct profiles of impairment in PD is limited. In this project, I will combine my experience with clinical neuroscience and my interest in machine learning to improve the predictive power of disease phenotypes in PD. I will do so by performing longitudinal analyses of neuroimaging, pathology and cognitive test data from a large group of PD patients and healthy controls, combined with machine learning techniques. The proposed work aims to yield individualized biomarkers of PD which will improve current best strategies for early diagnosis and treatment of PD.

Research Information:
We aim to study the spatiotemporal neural dynamics of the attentional blink with the EEG/fMRI data fusion method and computational modeling. The attentional blink occurs when the individual is unable to report the second target during a rapid serial visual presentation of images that contain two target images with a time distance of less than 500 msec. We will design an attentional blink paradigm and acquire EEG and fMRI data; moreover, we will develop a recurrent neural network to model the brain process in the attentional blink and perform various computational experiments on the model. Employing machine learning and deep neural network modeling, this research helps us to get an in-depth understanding of the brain's visual and attentional systems, which are two critical functions of the human brain. Additionally, by relating computational and theoretical aspects of attentional blink to practical aspects, this research will reveal new insights into consciousness and related deficits in clinical populations.

Research Information:
Patients diagnosed with schizophrenia display various patterns of symptoms and researchers have been trying to find the association between these symptoms and brain structural changes. One of these brain structural features is cortical gyrification, which indexes how the outer layer of the brain is folded. Previous studies have found abnormal cortical folding in schizophrenia patients, so we wonder — can this abnormal change in the brain structure predict the outcomes of schizophrenia patients (e.g. symptoms, cognitive abilities, and social functioning)? To answer this question, we will use cluster analysis, a machine learning tool to group schizophrenia patients with similar gyrification patterns together, and then track how their symptoms and cognitive abilities change over 2 years. If we can see there are profile differences between groups, it can shed light on the association between abnormal brain folding in the early stages of schizophrenia and certain symptoms. Understanding this association can better inform clinicians during diagnosis, predictions, and treatment.

Research Information:
I will study human brain's exceptional ability in recognizing objects under challenging conditions (such as occlusion or clutter) and model the visual neural mechanism using deep neural networks. The ventral visual pathway is involved in the visual recognition of objects through abundance feedforward and recurrent neural connections. The role of these connections in visual processing has been a long-standing question for neuroscience researchers and many studies have investigated the role of these neural connections in the visual cortex. We will go beyond the existing studies and investigate the computational role of neural recurrent and feedback processes, in particular, the algorithmic function of these processes.

Research Information:
What keeps the brain and body healthy throughout our lives? It has become increasingly clear that strong social ties are an important predictor of longevity. Conversely, loneliness is a strong risk factor for mental illness, poorer disease outcomes and mortality. Despite the importance of this topic, how social interaction improves our health is largely unknown. One theory is that social interaction reduces stress, and thereby reduces the stress hormone cortisol. Chronic exposure to cortisol is well-known to negatively affect mental and physical health. Research indicates that positive social interactions can suppress the brain's response to stress. This project will investigate the brain mechanisms through which positive social interaction reduces activation of the stress response and the stress hormone cortisol.

Research Information:
During primate evolution, the prefrontal cortex (PFC) has undergone a great expansion and is crucial to higher-order cognitive functions. The PFC is subject to substantial neuromodulation from all major ascending modulatory systems, and disrupted neuromodulation of PFC is a feature of numerous neuropsychiatric disorders including Alzheimer's disease, schizophrenia and attention deficit hyperactivity disorder. The impact of neuromodulation on PFC activity has been explored through extracellular electrophysiology, using both systemic and local pharmacology. From this, much is known about spiking activity (outputs) of PFC neurons and network activity through local field potentials, but little is understood about how neuromodulation affects synaptic activity and intrinsic mechanisms, which results in the spiking output in single PFC neurons during cognitive tasks. The development of intracellular recording methodology during active behaviour has yielded rich insights into how precise computations in cortical neurons are affected in sensory systems and in spatial navigation. This project seeks to revolutionize our understanding of the functioning of PFC circuits and the neuromodulation of cognition in awake-behaving NHP by combining neuropharmacological investigations with intracellular recordings. This will provide a comprehensive understanding of what computations PFC neurons perform in cognitive tasks and how these operations are affected in disease states.