Abigail Greene
About
Biography
Abigail is an MD/PhD candidate in the lab of Todd Constable. She received her PhD in 2021, with anticipated graduation in 2023. She received her A.B. with highest honors from Princeton in 2013, where she studied psychology in the lab of Jonathan Cohen, and received certificates in Quantitative and Computational Neuroscience and French. Following graduation, Abigail received a ReachOut56-81-06 fellowship to support a year working in the healthcare department of ProMujer Nicaragua.
Abigail's research focuses on the application of computational modeling and machine learning techniques to human neuroimaging data to reveal the neural bases of complex cognitive processes, traits, and clinical symptoms. She hopes to integrate this skillset with training as a psychiatrist to yield more precise understandings of the macroscale neural circuits underlying (dys)function. Outside of lab, Abigail worked with the Behavioral Health Department at Yale's student-run free clinic (HAVEN), which she directed from 2017-2018, and as a Pivotal Response Treatment clinician at the Child Study Center under the supervision of Dr. Pamela Ventola. She has presented her work at various conferences, including the Society for Neuroscience, Organization for Human Brain Mapping, and Flux Congress meetings, and has received various recognitions for her work, including an F1000 recommendation, the NIH Outstanding Scholars in Neuroscience award, and the YCCI Multidisciplinary Pre-doctoral Training Program in Translational Research Fellowship.
Education & Training
- AB
- Princeton University, Psychology (2013)
Research
Overview
Recent advances in human neuroimaging techniques have begun to offer insights into brain-behavior relationships; such work is exciting both because it sheds light on the neural underpinnings of complex traits, and because it offers the promise of predicting such traits on the basis of neural data, alone. Relevant patterns of neural activity are widely distributed across the brain, making whole-brain, data-driven, functional connectivity-based analyses well suited for their study.
Such analyses are commonly performed using data acquired at rest, an unconstrained state that may fail to capture the full range of individual differences in patterns of functional connectivity. Our research seeks to leverage such individual differences to predict individual traits, and has demonstrated that predictive models built using task-based functional connectivity (i.e., connectivity calculated using data acquired while subjects perform cognitive tasks [e.g., working memory and emotion identification tasks]) better predict phenotype than predictive models built using rest-based data. Thus, much like a stress test reveals variations in cardiac function not observable at rest, cognitive tasks reveal trait-relevant differences in functional connectivity. Our subsequent work has explored how tasks have this effect, and most recently applied this framework to a novel, clinically and demographically heterogeneous dataset, which includes task-based fMRI and neurocognitive measures, finding that brain-based models of phenotype recapitulate sample stereotypes; that is, one brain-phenotype relationship does not fit all. Together, this body of work seeks to develop and apply robust methods to relate patterns of brain activity to neurocognitive phenotypes with relevance to both health and disease.