Research

What psychological and neural mechanisms support human perception, learning, memory, language, problem-solving and social behaviors? Carnegie Mellon has been at the forefront in studying these questions for over five decades. Our research and training programs are strongly interdisciplinary, with an emphasis on the precise specification of the mechanisms underlying mental processes and behaviors, often incorporating neuroscientific data and computational models. Another hallmark of our programs is their integration across different levels of analysis: from structural and functional properties of neural circuits to fine-grained analysis of behavior, and from psychological models and large-scale simulations to intelligent artificial systems. We utilize a wide variety of methods to characterize behavior in normal and brain-damaged populations of children and adults, structural imaging, functional neuroimaging (including EEG, MEG, MRI and fNIRS), and computational modeling and simulation. We have helped to build science-based cognitive tutors, to understand how learning works across many different systems, to foster the creation of health neuroscience, and to put our findings into real-world practice in both the translational and commercial domains.

Computational neuroscience brings many ideas and tools associated with computation to the study of the nervous system. Major influences have come from the success of biophysical models of neural activity, the enduring appeal of the brain-as-computer metaphor, the increasing prominence of statistical and machine learning methods throughout science, and from modern advances in artificial intelligence. Here in Pittsburgh, we have an exceptionally large and vibrant community of neuroscientists and computational scientists who develop and/or apply cutting-edge computational methods in their work. We offer a Ph.D. through our Program in Neural Computation (PNC), an undergraduate minor in neural computation, year-long fellowships for CMU and University of Pittsburgh undergraduates, and a program of summer undergraduate research that draws students from across the U.S. Our research may be described, roughly, as falling into one or more of the following three broad categories: Modeling of Neurons and Neural Circuits, System and Cognitive Modeling, and Recording and Analysis of Network Activity. The computational neuroscience community here at CNBC consists of both faculty whose expertise is primarily computational and those who have expertise in experimental methods as well. Visit the Computational Neuroscience Faculty Directory for contact information. 
 

Neural technologies answer some of the most complex and long-standing questions about the brain. CMU researchers are developing next-generation technologies to study and model neural processes in both healthy and diseased states, and to assess and treat prevalent neurological conditions.

Neuroscience Institute researchers develop neurotechnologies with electrical, optical, ultrasonic and mechanical functionalities to interface with the brain across different scales of spatial and temporal resolution. Leading scientists and engineers advance neurotechnologies with a focus on Neuroengineering, NeuroAI and Neurorobotics.

• Neuroengineering - Neuroengineering research aims to develop novel tools and devices to study or treat the nervous system. Some focus areas within neuroengineering include designing and integrating brain-computer interfaces (BCI), improving assistive technologies and diagnosing or treating symptoms of neurological conditions with modulation techniques, enhanced neuroimaging and accurate biomechanical models. Progress in neuroengineering requires extensive collaboration between forward-thinking researchers, industry partners and clinicians to develop advanced technologies and translate discoveries into practical applications.
  • Current neuroengineering projects at CMU include:
    • Designing flexible and wearable electronics to record high-resolution muscle activity and electrical signals during movement for rehabilitation and surgical applications
    • Investigating neurostimulation techniques to improve motor control for stroke survivors and treat individuals suffering from chronic pain
    • Developing non-invasive, unbiased and accessible imaging tools for early diagnosis and real-time monitoring of neurological conditions such as epilepsy and traumatic brain injury
    • Creating a new class of nanoscale hybrid-materials for neuromodulation and biosensor applications
    • Designing BCI to investigate neural mechanisms of sensorimotor adaptations and skill acquisition for enhanced neural prosthetics
• NeuroAI - Artificial intelligence (AI) and its operating systems are designed to provide digital support, often performing complex tasks comparable to human behavior/intelligence. In scientific and clinical applications, AI is used for discovery, diagnosis and personalized technologies. At the Neuroscience Institute, experimentalists and computational researchers work together to develop neural-specific AI to explore and model neural processes and networks, efficiently record and analyze large neural data sets, and develop enhanced “human-like” digital systems and personalized assistive technologies.
  • Current neuroAI projects at CMU include:
    • Building machine learning algorithms to produce models of the brain (down to distinct features and connections of a single neuron, up to a cubic millimeter of cortex) with accurate automation, high resolution and widespread data accessibility 
      Improving data collection and analysis from functional neuroimaging techniques with novel machine learning solutions
    • Using AI models to encode how the brain represents (via activity and signal) language to formulate meaningful words and sentences
    • Innovating robust AI technology to decode and interpret human intention for movement through non-invasive BCIs
    • Encoding how brain areas have evolved to produce intelligent behavior using artificial neural networks that gradually improve and adapt over time
    • Developing enhanced AI tools to explain high-dimensional structure and time course of neural population activity for neural network reconstruction and fabrication of devices to interface with large neuron populations
• Neurorobotics - In collaboration with the College of Engineering and the Robotics Institute, students and faculty at the Neuroscience Institute are tackling complex neurohealth challenges with innovative robotics. Here, researchers utilize engineering and AI to build sophisticated neurorobotics with brain-inspired circuitry and architecture as models to study motor control and cognitive function. 

Systems neuroscience research at Carnegie Mellon is centered on understanding how the diversity of discrete neural cell types in the cerebral cortex and basal ganglia give rise to perception and behavior. Different types of neurons distinguished by patterns of gene expression and anatomical properties interact in stereotyped and hierarchical ways in order to generate complex behavior. New findings and sophisticated molecular tools for the identification, monitoring and control of neural activity in defined neural subtypes are revealing highly structured principles for information processing in the mammalian brain. Investigators use high-throughput methods to record neural firing, analyze cellular anatomy and characterize gene expression. In doing so, they are identifying specific computational properties carried out by molecularly-defined groups of neurons, as well as making discoveries about how they are changed by experience and disease. These findings are relevant not only to identifying new disease treatments, but also in the design and optimization of engineered networks for artificial intelligence and neural prosthetics.