Why Focus on Psychiatry?
There is well-founded disquiet regarding the classification of psychiatric disorders, as reflected in descriptive diagnostic systems such as the International Classification of Diseases (ICD) and the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association. This reflects a deeper concern, namely, that we lack an understanding of normal and deviant human cognition sufficient to bridge molecular, systems and phenomenological levels of description. We believe that a computational perspective, with its emphasis on formal models and quantitative analysis, is capable of building this bridge. As an example, consider the issue of helplessness, a common phenomenological feature in severe states of depression. From an experimental perspective we know that animals provided with rewards or punishments that their actions are incapable of influencing can be made helpless, and so, for instance, fail to explore when put into new environments. Computational insight derives from analysing helplessness as generalization between old and new of a statistical characteristic of environments, namely the control one expects to have of one’s own fate. This characteristic is captured in terms of a prior distribution over environments, a quantity that is under active examination in Bayesian approaches to cognitive science; furthermore model-based reinforcement learning is capable of incorporating such rich priors into decision-making. This analysis leads to a new view of environmental influences on reinforcement learning, new insight into the psychological and ultimately neural processes involved in depression, and even new assays of helplessness that can be assessed and tested in patient-based experiments or in prospective studies across the lifespan.
Why Focus on Behavioral Aging?
In most countries of the world, the older segment of the adult population is increasing in size, proportion, or both. Senescence operates continuously and cumulatively throughout adulthood, altering the neurochemistry, anatomy, and functional dynamics of the aging brain. There is a pressing need to identify and strengthen the environmental conditions and biological mechanisms that favor desirable developmental trajectories in later adulthood and old age. Maintaining cognitive abilities and postponing or preventing pathologies leading to dementia are key aims in this endeavor. Existing evidence is descriptive rather than explanatory, focusing on documenting the large heterogeneity of cognitive abilities and developmental trajectories in old age. To delineate the causes of this heterogeneity, and thus promote desirable outcomes, we need a systems perspective that connects changes in vasculature, structure, neurotransmission, and metabolism of the brain to changes in its function, and to changes in behavior. Combining advanced structural and functional imaging methods with statistical and computational methods aiming at the classification and prediction of patterns of individual change will help to identify the sources, including common genetic polymorphisms, contributing to the increasing heterogeneity of cognitive performance with advancing adult age. Such studies will help to delineate the relative importance of maintenance, restoration, selection, and compensation as mechanisms of preserved cognitive competence in old age, and suggest suitable time windows and targets for prevention and remediation.
Why Combine the Two?
Understanding and ameliorating psychiatric disease and cognitive aging have important commonalities, and, in some cases, are associated with overlapping aetiologies. Both endeavors are of formidable complexity and must be approximated by focusing on key players at the neural level, such as the neurovascular unit, white matter, dopaminergic neurotransmission, variability of neural signalling, and striatal-cortical circuits, and on key players in behavior, such as working memory, learning, and decision-making. Computational models connect structure to function and behavior. By applying these methods to experimental and, critically, longitudinal data, we will be able to derive testable predictions about key determinants of psychiatric syndromes and cognitive aging, and provide a more solid empirical and theoretical basis for the attenuation and amelioration of psychopathology and cognitive decline through personalized interventions.