Individuals who are prone to experiencing high levels of anxiety and depression often exhibit dysfunctional behavior. For example, anxious individuals often avoid situations that have even the slightest chance of a highly negative outcome (e.g. a plane crash), and depressed individuals often show a reduced pursuit of activities that most people find enjoyable. Progress can be made in understanding dysfunctional behavior by using formal, mathematical frameworks of decision making, which break down behavior into its computational components, and in which we can start to pinpoint the specific abnormalities associated with anxiety and depression. Chapter 1.2 and 1.3 review prior literature, highlighting some of the computations that seem to be altered, such as the overestimation for the probability that rare, extremely negative events will occur. Chapter 2 and Chapter 3 empirically examine behavior in situations that require individuals to accurately estimate the probability that an outcome will (or will not) occur as a result of their actions. In a task where individuals have to estimate action-outcome probabilities by trial-and-error (Chapter 2), individuals with high overall levels of anxiety and depression show a reduced ability to align the rate at which they learn to the rate of change in the environment (i.e. the level of volatility). In a task where individuals have to choose between options that have known (risky) versus unknown (ambiguous) probabilities (Chapter 3), individuals who have high levels of physiological anxiety tend to avoid the ambiguous options more than other individuals, as information is removed about those probabilities. On the other hand, individuals who are prone to experiencing mania are more likely to make the opposite choice, seeking ambiguity, when the outcomes are rewarding. Chapter 4 examines possible sources for dysfunctional beliefs, as opposed to behaviors. In a hypothetical vocational setting where individuals estimate their rank relative to others, individuals with high levels of anhedonia-related symptoms show initial beliefs that are more negative relative to the beliefs of others. Individuals with high levels of anxiety, on the other hand, show negatively biased updating of those beliefs in response to unbiased information. Chapter 5 summarizes the empirical findings and discusses more broadly how anxiety and depression seem to impact behavior (and its underlying computations) in uncertain situations.

Emotions are stereotyped responses to situations of high survival value. Recognition of these high survival value situations is a necessary precondition to emotional elicitation. As vision is the primary human sensory modality, visual perception of an emotional situation is often responsible for the elicitation of emotion in humans. This raises the question: how does the human brain extract the emotional content from patterns of light on the retina in order to initiate appropriate behavioral responses that are adaptive for survival? And how are these emotional scenes differentially processed from non-emotional scenes? In order to address these questions I conducted a series of experiments using techniques and theory from a broad range of disciplines including cognitive neuroscience, machine learning, psychophysics, and affective science. I motivate, describe, and interpret these experiments in the five chapters of this dissertation.

Chapter 1 contains a brief review of the previous scientific findings that motivated this dissertation. Functional magnetic resonance imaging (fMRI) is a non-invasive neuroimaging modality which allows researchers to record brain activity with high spatial resolution. Traditional techniques used in the analysis of fMRI data have recently been complimented by computational machine learning techniques, giving neuroscientists powerful new tools to infer the brain’s mechanisms and representations. Voxel-wise modeling (VWM) is one such method. Chapter 2 describes a fMRI experiment using a large corpus of naturalistic emotional images and VWM analysis to model brain representations of a combined semantic and emotional feature space. Principal components analysis (PCA) of voxel tuning was then used to uncover the primary dimensions of representation within occipital-temporal cortex (OTC). Alongside animacy, the valence and arousal of animate stimuli are primary dimensions of OTC tuning. Furthermore, this tuning is better able to predict the appropriate behavioral response to the images that subjects viewed than are the semantic and emotional image features used to model the fMRI data. These findings suggest that OTC representations of naturalistic emotional images may be used by other brain regions to elicit appropriate behavioral responses to situations of high survival value depicted in these emotional scenes.

In order to address several theoretical limitations of the study described in chapter 2, I conducted two literature reviews in Chapter 3. The first reviews studies on congenitally blind subjects which suggest that OTC contains supramodal semantic representations that may be idiosyncratically tuned to support appropriate behavior responses. The second reviews the literature on attention and perception of emotional images and the neural mechanisms that subserve these processes. Additionally, Chapter 3 describes seven fMRI analyses of the data from chapter 2 that address empirical limitations of chapter 2. These analyses include a representational similarity analysis (RSA), univariate SPM analysis, variance partitioning of the VWM model in chapter 2, and several control analyses. The findings from these analyses further boost our claims from chapter 2 that alongside animacy, the valence and arousal of animate stimuli are represented within OTC.

Often situations of high survival value occur quickly, or are only perceived briefly, and thus the human visual system has little information from which action can be taken. Turning from cortical to behavioral responses from naturalistic emotional images, in chapter 4 I describe an experiment using ultra-rapid image presentation to explore the limits of human performance in semantic and emotional valence categorization during a brief glance. Using stimuli and categorization tasks from a broader range of semantic and valence categories than had been done in previous studies, along with controls for observed response bias, I found that humans can accurately categorize both the semantic category (animal, human, object, or building) and emotional valence (negative, neutral or positive) of an image when presented for as little as 17ms and backwards masked. Furthermore, when the image depicted an emotionally negative scene, semantic performance was significantly worse than that for images depicting neutral or positive scenes. I also found that only when subjects successfully categorized the semantic category of an image was valence categorization also above chance across several valence by semantic category conditions. The converse was not true however. This suggests that the semantic information of an image must first be extracted before the emotional information can be extracted. This finding supports the cognitive primacy hypothesis as opposed to the affective primacy hypothesis, two competing hypotheses of visual emotional processing in the field of affective science. Finally, in chapter 5, I interpret the results of the experiments described in chapters 2-4 taken as a whole, and offer closing remarks.

Computational models of cognitive processes have provided deep insights into specific mechanisms of human learning and decision-making. A natural corollary of research into the typical functioning of such mechanisms is to investigate how mental disorders might cause impairment of the underlying algorithmic processes. This line of research, frequently referred to as computational psychiatry, seeks to contribute to more personalized diagnoses and treatments of mental disorders by characterizing behavioral and computational profiles of psychopathology. In the current dissertation we focus on the role of latent dimensions of internalizing psychopathology in computational processes of learning and decision-making to provide insight into dissociable dimensional phenotypes. Chapter 1 of the dissertation provides a primer to computational characterizations of learning and decision-making, including recent and foundational research results in computational psychiatry which motivate the subsequent original research questions. In Chapter 2, we present an investigation of effortful decision-making in reward pursuit and threat avoidance. We use a detailed dimensional profile of depressive and anxious symptoms to characterize unique mechanistic impairments across frequently comorbid symptoms, with individual symptoms of depression relating to specific differences in effortful reward pursuit processes, while a dimensional characterization of anxiety relates to multiple mechanistic differences in effortful threat avoidance. In Chapter 3, we extend the space of psychopathological inquiry to include subclinical levels of mania in an investigation of decision-making under ambiguity in reward pursuit, loss avoidance, and threat avoidance paradigms. Here, individual differences in loss avoidance decisions under ambiguity show opposing effects of anhedonic depression and physiological anxiety in risk sensitivity, with depression showing increased risk sensitivity and anxiety showing reduced nonlinear risk valuation. Ambiguous decisions in threat avoidance reveal a significant dissociation between anhedonic depression and hypomania, with higher levels of hypomania associated with significantly less risk sensitivity in threat avoidance than anhedonic depression. We also find suggestive, although not concrete, evidence that sensitivity to threat magnitude may relate to comorbid variance between anxiety and depression, and we recommend additional work in this area of investigation. In Chapter 4, we present a detailed investigation of dimensional characterizations of anxiety in reinforcement learning impairment, using an experimental paradigm which differentially manipulates the relative load of working memory versus learning systems across task. Here we find that physiological, but not cognitive, anxiety is related to significant learning impairments which are contributed to both by slower learning processes and by increased rate of working memory decay. These results may have implications for a wide variety of extant reinforcement learning modeling studies in computational psychiatry by emphasizing the importance of an algorithmic account of supporting working memory processes in such investigations. Chapter 5 provides a discussion of the results in the framework of dissociating unique profiles of dimensions of internalizing psychopathology. As additional insights into the relationships of latent dimensions of psychopathology with algorithms of cognition are gained, the field as a whole is enabled to approach the idea of replicable computational phenotypes of psychopathology. The aim of this dissertation is not to operationalize such phenotypes at present, but rather to contribute to the growing body of knowledge within the field to enable and inspire more detailed investigations and replications of the findings contained herein.