Abstract:
|
Promising laboratory-based findings often fail to translate to clinical populations, leading to failures to predict and treat psychopathology based on promising biological targets. This failure may be due to foundational differences in how clinical populations are defined, where laboratory research focuses on direct behavioral, physiological, or neurobiological functioning while clinical research focuses on DSM-based clinical definitions. To bridge this translational gap requires methods to define and predict clinical populations based on basic dimensions of behavior, physiology, and neurobiology in a manner that is scalable and transportable beyond the laboratory. Such an approach allows for the identification of risk and targets for remediation individualized to the patient’s underlying deficit (personalized medicine). We will present new approaches to define and predict clinical functioning and course based on computational methods in machine learning and artificial intelligence that are free from traditional diagnostic mile-markers along with approaches to develop tools and approaches to provide new, data driven, definitions and predictive approaches to health and pathology.
|