Center of Alcohol Studies
Rutgers University–New Brunswick
Research on implicit processes has revealed multiple shortcomings of awareness categorization tools, which are most often based on the failure to score significantly above chance level. Moreover, post-hoc awareness categorizations result in regression to the mean effects, by which aware participants are wrongly categorized as unaware. Using Bayes factors to obtain sensitive evidence for participants’ lack of knowledge could prevent regression to the mean effects. Here I present the reliability analysis of a novel Bayesian awareness categorization procedure, comparing it to traditional approaches based on t-tests.
Following a reward conditioning task, we divided participants in different groups (Aware, Unaware, Insensitive) depending on the results of individual Bayesian analysis ran on an awareness measure. We examined the cumulative awareness scores of each group, and found that Unaware participants did score below chance level, contrary to the predictions of regression to the mean effect. This was further confirmed using a resampling procedure with multiple iterations. Conversely, when categorizing participants using t-tests, Unaware participants scored significantly above chance level on the resampling procedure, showing regression to the mean effects.
The results show that using Bayes factors instead of t-tests as a categorization tool it is possible to avoid regression to the mean effects. The reliability of the Bayesian awareness categorization procedure strengthens previous evidence for implicit reward conditioning. I will discuss how Bayes factor can reduce measurement errors to provide a reliable post-hoc awareness categorization tool, improving the quality of research on implicit processes. The toolbox necessary to perform the categorization procedure is available to be incorporated in future experiments.
Lifespan Brain Institute
Children’s Hospital of Philadelphia and UPenn Medicine
Department of Psychiatry
UCL Max Planck Centre for Computational Psychiatry and Ageing Research
University College London
The burden of depression is to no small part due to its chronic or recurring
nature. As such, the maintenance of any treatment gains is of paramount
importance. A key step in this process is the decision to discontinue
antidepressant medication. However, at present there are no predictors to
indicate who can safely discontinue medication. The AIDA study recruited
123 patients who had remitted on antidepressant medication and were intent
on discontinuing their medication. Patients were randomized into two groups.
Both groups underwent two extensive assessments involving clinical,
behavioural, imaging and biochemical assessments, but one group was tested
before and after discontinuing antidepressants, while the other was tested
twice before discontinuation. Patients were followed up for 6 months to
monitor for relapses.
57 healthy, never-depressed matched controls were recruited. Of 104
patients who completed at least one assessment, 84 completed the study, with
34 relapsing during the follow-up. Amongst standard clinical variables, only
treatment by non-specialists was robustly associated with relapse (p=0.005),
but did not predict relapse out-of-sample. In contrast, several behavioural
(effort-related), psychological (brooding rumination, neuroticism) and
imaging (EEG alpha asymmetry) variables had predictive power (all p<.05),
while others were affected by discontinuation or distinguished remitted
patients from healthy controls. Relapse after antidepressant discontinuation
can be predicted by a number of variables. A combination of these may reach
an accuracy sufficient to guide clinical decision-making.
Christen Crosta and Bonnie L. Firestein
Department of Cell Biology and Neuroscience
Proteins Associated with Schizophrenia in Epithelial Cheek Cells and Relationship to Optical Coherence Tomography Data and Cognitive Symptoms
Natural environments are feature-rich and only a subset of these features are considered to predict action-outcome associations. To enable accurate action-outcome predictions a decision-maker is faced with a challenge, namely that only a portion of the information in the environment is predictive of a desired outcome. Here, we highlight the tendency of individuals to assign credit to outcome-irrelevant task representations. We demonstrate that value is assigned to these representations in a model-free and state-independent manner. We further show the association between these low-level value associations and a more sophisticated model-based system, and propose how model-free representations might be regulated according to a model of the environment. Finally, we suggest that a deficit in the regulation of outcome-irrelevant model-free associations might lead to behavioral abnormalities such as compulsive behavior.
Department of Psychiatry
University of Pennsylvania School of Medicine
Increasingly, large-scale studies of brain development incorporate complex imaging data alongside diverse clinical and cognitive measures. However, it remains a challenge to integrate these high-dimensional data, and generate measures tailored to the individual. Here I will focus on recent work illustrates how data integration and personalization can inform our understanding of normal and abnormal brain development. Specifically, I will demonstrate how machine learning techniques can help identify links between dimensions of psychopathology in youth and abnormalities within functional brain networks. Furthermore, I will present unpublished work that uses a combination of functional imaging data and machine learning techniques to identify personalized networks in the developing brain. Together, such approaches may evolve to help produce biomarkers of risk and resilience in youth.
Mood Brain and Development Unit
National Institute of Mental Health
Implementing analytic tools from engineering can bring a great promise to the study of human dynamics. In this talk I will show how a closed-loop control strategy can be a powerful tool for controlling and characterizing mood. I will present the Mood-Machine-Interface, an efficient mood modification algorithm, that can shift mood by adjusting reward values according to individual mood sensitivity. This method ensures that the created positive and negative environments are potent across individuals and across time. The paradigm can be used to generate substantial mood changes in healthy, depressed, adult and adolescent individuals. And moreover, by creating a non-random environment, it enabled us to model how mood changes are affected by previous events. This model shows that early experiences carry a strong influence on mood, which is mediated by neural signals in the ACC, thus providing a neuro-computational account for mood regulation. Overall, this paradigm demonstrates how closing the loop can open the door to understanding mood dynamics and to developing individualized treatments of mood disorders.
Michael W. Cole
Center for Molecular and Behavioral Neuroscience (CMBN)
Rutgers University – Newark
Understanding neurocognitive computations – such as those disrupted in mental disorders – will require not just localizing cognitive information distributed throughout the brain but also determining how that information got there. Brain connectivity clearly has something to do with it, and decades of “connectionist” (and recent “deep learning”) theory suggests connectivity patterns specify distributed neural computations. I will share my laboratory’s efforts to map the human brain’s functional network organization and to determine how that organization shapes distributed cognitive processes and mental health. This has involved identification of “flexible hubs” – cortical regions that adaptively shift their connectivity to implement instructed task procedures. We have recently linked flexible hubs to the regulation of mental health. Also central to these efforts is the estimation of activity flow – the movement of evoked activity through brain network connections. Estimating activity flow quantifies the likely contribution of a network organization (such as one estimated using resting-state functional connectivity) to function-specific activity patterns (such as fMRI responses to cognitive events). I will cover application of activity flow mapping to predict whole-brain activation patterns across a variety of tasks, such as tasks involving working memory and rapid instructed task learning, as well as in the context of pre-clinical Alzheimer’s disease. These developments promise to better integrate theoretical/computational neuroscience and empirical neuroscience with mutual benefits across both fields.
Judith Mildner and Elyssa Barrick
Department of Psychology
In our daily lives, thoughts often unfold spontaneously and freely, for instance during dreaming, mind wandering, daydreaming, or creative ideation. One theory of spontaneous thought suggests that these thoughts arise from a dynamic exploration of semantic and episodic memory that is influenced by the thinker’s current state. While creativity is typically considered a trait, like intelligence, this view of spontaneous thought suggests that a key component of creativity, creative ideation, is in fact more state-like. In this project, we aim to investigate the changing dynamics of spontaneous thought and creative ideation by examining thought in patients with bipolar disorder. How do patterns of spontaneous thought change between manic and euthymic states in these patients? And how do these changes relate to creative thought?
Princeton Neuroscience Institute
Dynamic expectations: The interplay of timing and reward prediction during learning