To understand how the brain/mind works, we employ computational simulations, model building, and advanced statistical data analyses. Our research ranges from computational neuroscience, to computational models of perceptual and motor inferences, to computational models of higher-level cognitive processes, such as decision-making, language acquisition/processing, and numerical reasoning.
For example, brain and cognitive sciences (BCS) researchers have developed computational theories and models of how:
- Neural systems conduct inferences about the world and how specific properties of neural populations support this ability (neural noise structure; population codes; neurally implemented statistical inference)
- The brain integrates multiple sources of information, such as when recognizing objects from touch when we have only seen them before (cue combination; cross-modal inference; multi-sensory integration)
- We generalize from previous experience to novel situations (similarity-based and hierarchical inference)
- We understand speech despite radical variations in the physical sounds that reach our ears (word recognition, speech perception and adaptation)
- We acquire words, including the meaning of highly abstract words like “the”, without explicit instruction (segmentation; lexical and syntactic learning)
The University of Rochester’s Center for Integrated Computing provides access to an IBM Blue Gene/Q system, a Blue Hive Linux cluster, and other distributed computing resources, providing 420 teraFLOPS and 2 petabytes of storage. Additional large-scale computing resources specifically tailored to our research demands are available through individual labs and through our collaborations with computer science. These resources facilitate the use of massive data sets, statistical simulations, and large-scale simulations of neural circuits.
BCS researchers are well known for innovative and cutting-edge use of behavioral methods. Our labs offer access to a broad range of behavioral paradigms and equipment for use with both humans and animals. This includes numerous head-mounted and stand-alone eye-trackers suitable for research on infants, kids, adults, and research animals, as well as infrared motion tracking systems to quantify visuomotor control.
To study behavior outside of the laboratory in more natural situations, we make use of wearable eye-trackers and other immersive technologies, including automatically activated wearable recording devices and remote data collection applications on tablets or cell phones. Behavioral studies beyond the traditional lab setting also involve field work (e.g., studies on indigenous populations) and the extensive use of web-based experiments, including large-scale studies on language learning, speech perception and production. Crowdsourcing platforms allow collection of tens of thousands of data points over very short time periods.
Students and faculty in BCS also have access to virtual reality environments, including a 180-degree field of view virtual reality (VR) system (pictured right) and force phantoms that allow simulation of tactile sensation. These assets facilitate studies of decision-making, motor planning, and perceptual inference. For example, VR environments allow us to study how the reliability of different sensory cues, such as tactile and visual information, affects their relative weighting in higher-level perceptual inference processes, such as recognizing and manipulating objects. Similarly, speech perturbation setups allow us to study how speakers monitor and adjust their pronunciations when they (are tricked into perceiving themselves to) deviate from their target productions.
BCS researchers use neuroimaging methods to understand complex cognitive processes in terms of their underlying neural mechanisms. Imaging modalities include BOLD MRI, DTI, high resolution anatomical MRI, MRI spectroscopy, TMS, tDCS, EEG, and NIRS. A range of sophisticated functional image analysis techniques are developed and implemented including machine learning and multivariate approaches, model-based fMRI, graphical modeling, functional connectivity, and inter-subject correlations.
Structural analysis approaches include voxel-based morphometry and cortical thickness analysis, fiber tractography, and lesion overlap analysis. Investigations employ these complementary tools to understand the neural bases of conceptual knowledge, perception and motor planning, language, decision-making, learning, and memory. Current research uses brain imaging technologies to study cognitive phenomena in human infants, children, and adults as well as neuropsychological populations. Student research projects in cognitive neuroscience are shaped by the department's strong emphasis on combining behavioral, computational, and neural investigations.
A research-dedicated 3T MRI machine (pictured right) is operated by the Rochester Center for Brain Imaging – a multi-department neuroimaging research center that is closely affiliated with the Department of Brain and Cognitive Sciences and also includes faculty from engineering, physics, computer science, and medicine. Individual investigators also have access to shared EEG, TMS, and BrainSight instrumentation.
Students and faculty in BCS employ animal models to uncover fundamental mechanisms by which neural circuits in the brain give rise to basic aspects of mental function, including perception, decision-making, attention, working memory, learning, motor planning, and judgments of value. BCS research laboratories are equipped for studies with non-human primates, birds, and rodents. Modern equipment allows recording of action potentials from populations of neurons in alert trained animals, along with simultaneous monitoring of eye and head movements.
Virtual-reality systems allow animals to forage in or navigate through artificial environments which can be precisely controlled. MRI and BrainSight facilities are available for structural and functional mapping of brain regions in laboratory animals.
BCS faculty are leading studies that seek to define the neural underpinnings of value-based decision making, self control, visual and vestibular perception, multisensory integration, visual object recognition, and attention. Simultaneous behavioral and physiological studies are combined with computational modeling to achieve new insights into the processes by which patterns of distributed neural activity lead to complex behaviors.
The study of neuropsychological populations can yield important insights into normal brain function and how it is compromised by injury, disease, or aging. BCS students and faculty work with various groups of patients, including individuals with brain damage, stroke, autism, and low-vision, as well as unique populations such as individuals with synesthesia.
These studies inform models of normal brain processing, help constrain localization of function to specific brain regions, and, in some cases, provide causal evidence for links between brain structure and brain function. Additionally, studies with special populations have the potential to translate into future treatments that may lead to improvements in cognition, perception and language.
BCS researchers develop cognitive and perceptual interventions and use a variety of experimental approaches to track outcomes and progress of both cognitive and clinical treatments.