Palm Health Foundation Computational Brain Health Graduate Fellows

Brain, Neuroscience, Brain Health

By gisele galoustian | 11/30/2023

�鶹Ů�� ’s Stiles-Nicholson Brain Institute has announced the second round of awardees of the “Computational Brain Science and Health Graduate Fellowships.” A generous gift of $1 million from the Palm Health Foundation () in 2022, awarded through its Brain Health Innovation Fund, supports new technologies, treatments, resources and educational tools to advance brain health in the community. Five �鶹Ů�� Ph.D. students, whose work embraces computational neuroscience, have been selected to receive the fellowships.

“Through the generosity and continued support from the Palm Health Foundation, talented young �鶹Ů�� scientists will have the opportunity to advance their important discoveries that benefit from computational approaches,” said Randy D. Blakely, Ph.D., executive director, �鶹Ů�� Stiles-Nicholson Brain Institute, the David J.S. Nicholson Distinguished Professor in Neuroscience, and a professor in the Department of Biomedical Science within �鶹Ů��’s . “Their novel research is aimed at data-intensive investigations underlying autism, Huntington’s disease, neurostimulation, a secure encryption method for large medical imaging file formats and elucidating the integral role the prefrontal cortex plays in complex behaviors.”

The students will be collaborating with �鶹Ů�� faculty on advanced research targeted at understanding the underpinnings of some of the most complex brain disorders to ultimately develop innovative methods, treatments and therapies.

“The five recipients of the Palm Health Foundation Computational Brain Health Graduate Fellowships were selected because of the high quality of their research and unwavering dedication to improving quality of life,” said , Ph.D., associate professor, �鶹Ů�� , and director of research development and diversity, �鶹Ů�� Stiles-Nicholson Brain Institution, who spearheaded the selection process. “Graduate students often juggle multiple priorities from teaching to seminars to working in the lab. By funding their projects, we are helping to alleviate these challenges and providing them with the opportunity to focus on their important research.”

The recipients of the PHF Computational Brain Health Graduate Fellowships are:

Lindsey Riera-Gomez, “Neural Synchrony During Parent-infant Interactions in Infants At-risk for Autism” (Mentor: , Ph.D., professor of psychology, �鶹Ů�� Charles E. Schmidt College of Science):

The goal of this project is to identify early markers for autism spectrum disorders (ASD) during infancy, which may aid clinicians in providing earlier diagnoses and treatment at a time when the brain is most plastic. Some have speculated that interpersonal synchrony (coordinated behaviors, movements, and neural activation) between an infant and their parent is an important developmental milestone that is predictive of social brain development, empathy, and symbol-use, and may be disrupted in infants at-familial-risk for ASD. Functional Near-Infrared Spectroscopy (fNIRS) is a neuroimaging tool that is ideal for use in both young and clinical populations. Given that ASD is a neurological disorder that disrupts social functioning, an fNIRS hyperscanning (simultaneous brain scanning of both parent and infant during a social interaction) approach could help to identify the neural underpinnings of social deficits in ASD during naturalistic face-to-face interactions. An fNIRS neuroimaging approach could identify if the level of interpersonal synchrony that infants share with their parent is associated with their familial risk for ASD.

Gianna Cannestro, “Computational Approaches to Optimize Data Analysis for a Huntington’s Disease Model” (Mentor: Jianning “Jenny” Wei , Ph.D., associate professor of biomedical science, �鶹Ů�� Schmidt College of Medicine):

Huntington’s disease (HD) is a heritable, terminal, neurodegenerative disease with no known treatment or cure resulting in emotional, cognitive, and motor dysfunction. In healthy individuals, the huntingtin gene has 10 to 34 CAG repeats, but more than 40 repeats can result in the disease state. However, mice, the primary HD animal model, require a higher number of repeats than humans do to exhibit characteristic HD phenotypes and therefore, may not be the most suitable model to study HD for therapeutic interventions. The utilization of human-derived induced pluripotent stem cells (iPSCs) could minimize interspecies differences and represent a model more like the in vitro human disease state. Cannestro and collaborators plan to utilize methods of signal processing to tease apart electrical recordings into functional and connective differences between healthy and HD cells. They will determine potential differences in electrical activity of individual cells, the connective characteristics as the state of intercellular connections, and the resulting network activity. Specifically, this project focuses on optimizing algorithms and developing a tailored computational analysis pipeline to handle and process the large amount of data recorded from multi electrode array experiments. Identifying electrical and connective differences between HD and healthy cells can aid in the development of diagnostic tools and the identification of potential cellular and subcellular mechanisms being affected by the disease with better fidelity to the human disease state. It may also provide a framework for the development of disease detection and drug screening methodology. This neuronal activity-based functional platform using human iPSC-derived neurons would greatly enhance the drug discovery process for hundreds of thousands of patients in the United States alone.

Joseph McKinley, “Neural Entrainment: A New Complexity Science Paradigm for Healing the Brain” (Mentor: , Ph.D., associate professor of physics, �鶹Ů�� Charles E. Schmidt College of Science):

Substantial effort has been devoted to investigating the health applications of neurostimulation, a therapeutic intervention whereby electromagnetic signals are used to disrupt pathological patterns of neural activity associated with brain disease and induce new healthier patterns. However, while neurostimulation has been shown to effectively treat many neuropsychiatric disorders, the underlying mechanisms of action remain poorly understood, and as such, neurostimulation’s full utility as a health intervention has yet to be realized. McKinley aims to apply these insights to develop a nonlinear dynamical theory of neurostimulation, with the goal of informing precise and individualized neurostimulation treatment. By generalizing to the case of time-dependent neurostimulations, the theory will elucidate the underlying dynamical principles of neurostimulation, maximizing its safety and efficacy as a health intervention as well as minimizing side effects and uncertainty of treatment outcomes, honing a new tool for healing the brain. This research has implications for a wide range of brain-based diseases, including neurological diseases such as Parkinson’s and Alzheimer’s, epilepsy, tinnitus, and chronic pain, and psychiatric disorders including major depression, obsessive compulsive disorder, generalized anxiety, post-traumatic stress disorder, bipolar disorder, and psychosis. Resulting work from this research will give clinical neuroscientists the precision to prescribe treatments based on the needs of individual patients according to their unique neural makeup and specific pathologies and will allow tailored treatments that maximize efficacy and safety while minimizing side effects and risk, improving treatment outcomes, and setting the foundation for future neuroscience research investigating brain health.

Jennifer Giordano, “Safeguarding the Brain: Secure Neuroimaging Data Encryption for AI-Driven Brain Analysis” (Mentor: , Ph.D., assistant professor of mathematical sciences, �鶹Ů�� Charles E. Schmidt College of Science):

The future of medicine is shifting to cloud computing, where health care providers can access patient information from anywhere at any time. This transformative shift will facilitate rapid and efficient treatment decisions, which can be crucial in emergency situations. As the fields of AI and neuroscience advance, the development of methods to interpret brain activity becomes increasingly imminent. Recent studies have already demonstrated the feasibility of such interpretation for specific brain regions. Decoding patterns of neural activity will unlock insights into the underlying mechanisms of cognitive function and behavior, potentially revolutionizing medical treatments. However, these developments also present challenges to HIPAA regulation and raise additional privacy concerns for intellectual property derived from this information. Moreover, a basic goal of integrating AI into the medical field is to train agents to identify abnormal brain scans. To do this, datasets must be expanded by several orders of magnitude, because medical diagnoses often are highly individualized with significant variability in symptom characteristics. Ensuring data security is crucial in protecting patients’ privacy and encouraging participation in neuroimaging studies. Ultimately, the integration of AI and cloud computing has the potential to revolutionize various aspects of health care. Giordano’s research will help address a vital step toward this goal: the creation of a secure encryption method capable of handling large medical imaging file formats while preserving the content and patient privacy.

Ryan Gallagher, “The Influence of Presentation Order and Duration on Learning Task Structure” (Mentor: , Ph.D., assistant professor of psychology, �鶹Ů�� Charles E. Schmidt College of Science):

The prefrontal cortex (PFC) plays an integral role in complex behaviors that enable individuals to adapt to ever changing environmental demands. Computational models of PFC have been developed to explain the neural mechanisms instantiating these processes. The Hierarchical Error Representation (HER) model proposes a unifying computational account of PFC function that can replicate activity observed in single-unit functional magnetic resonance imaging (fMRI) and EEF studies, while also producing behavioral evidence of more complex structured learning. Gallagher’s research seeks to advance understanding of the neural mechanisms of PFC and its role in behavior. To accomplish this, he will employ a combination of behavioral experiments and computational modeling techniques. He has designed a series of experiments that manipulate various aspects of stimulus presentation (presentation order and duration on a learning task) to investigate how temporal and structural abstractions interact to govern how multiple sources of information are integrated. Findings from this research will provide information about the neural mechanisms that govern how multiple sources of information are integrated, which can then be used to refine theoretical models of the PFC.

“The research being conducted by some of the brightest minds in the nation will greatly contribute to our knowledge of brain function and brain health and will enhance the well-being of our communities in Palm Beach County and beyond,” said Patrick McNamara, president and CEO of PHF. “We are excited and proud to support these five �鶹Ů�� graduate students selected to receive the Computational Brain Science and Health Graduate Fellowships, who are exploring unchartered territories in neuroscience and advancing scientific research to new levels.”

-�鶹Ů��-