Active interoceptive inference and the emotional brain.

Seth AK, Friston KJ. · 2016

Read Active interoceptive inference and the emotional brain. by Seth AK, Friston KJ. (2016) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5062097.txt.

Introduction

Anil K. Seth and Karl J. Friston review interoceptive inference in Philosophical Transactions of the Royal Society B in 2016. Interoception is the sensing of the body's inner state: heartbeat, breath, gut, and blood chemistry. The authors recast it inside predictive processing, the view that the brain constantly predicts its inputs and corrects the errors.

On this account, the brain does not passively wait for body signals. It sends predictions down to the reflexes that control the body, and those reflexes act to make the predictions come true, holding inner conditions steady or shifting them ahead of demand. How much trust each signal gets, its precision, may be set by neuromodulators. The frame reshapes thinking about the emotional brain and the sense of being an embodied self. It also reshapes thinking on disorders from autism to depression, all of which the authors survey as test cases for the theory.

Catalog record

CREATORSeth AK, Friston KJ.
YEAR2016
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5062097

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Active interoceptive inference and the emotional brain By Anil K. Seth, Karl J. Friston Published in: Philosophical Transactions of the Royal Society B: Biological Sciences DOI: 10.1098/rstb.2016.0007 ABSTRACT We review a recent shift in conceptions of interoception and its relationship to hierarchical inference in the brain. The notion of interoceptive inference means that bodily states are regulated by autonomic reflexes that are enslaved by descending predictions from deep generative models of our internal and external milieu. This re-conceptualization illuminates several issues in cognitive and clinical neuroscience with implications for experiences of selfhood and emotion. We first contextualize interoception in terms of active (Bayesian) inference in the brain, highlighting its enactivist (embodied) aspects. We then consider the key role of uncertainty or precision and how this might translate into neuromodulation. We next examine the implications for understanding the functional anatomy of the emotional brain, surveying recent observations on agranular cortex. Finally, we turn to theoretical issues, namely, the role of interoception in shaping a sense of embodied self and feelings. We will draw links between physiological homoeostasis and allostasis, early cybernetic ideas of predictive control and hierarchical generative models in predictive processing. The explanatory scope of interoceptive inference ranges from explanations for autism and depression, through to consciousness. We offer a brief survey of these exciting developments. This article is part of the themed issue ‘Interoception beyond homeostasis: affect, cognition and mental health’. INTRODUCTION Recent years have seen the emergence of a framework within cognitive neuroscience that offers exactly the right set of concepts to talk about the body and mind in terms of beliefs about the body (and oneself). On this view, the brain is not an elaborate stimulus-response link but a statistical organ that actively generates explanations for the stimuli it encounters—in terms of hypotheses that are tested against sensory evidence. This perspective can be traced back to Helmholtzian formulations of unconscious inference [1]. Over the last few years, the underlying idea has been formalized to cover deep or hierarchical Bayesian inference–about the hidden causes of our sensations—and how these inferences induce beliefs and behaviour [2–7]. ‘Explanations’, ‘hypotheses’ and ‘beliefs’ should in this context be understood not as consciously held mental states, but as neuronally encoded probability distributions (i.e. Bayesian beliefs) over the hidden causes of sensory signals. The biophysical encoding of these ‘beliefs’ is, technically, in terms of sufficient statistics like the mean or expectation of a distribution. In the last few years, ‘Bayesian brain’ ideas have been applied in the context of interoception (figure 1), which refers to the perception and integration of autonomic, hormonal, visceral and immunological signals [8,9]—or more informally as the sense of the body ‘from within’. On some of these views [7,10], emotional experience and experiences of embodied selfhood emerge from top-down inference on the (multimodal) causes of interoceptive afferents, generalizing so-called two-factor or evaluative theories of emotion and cognition [11]. A first implication of these proposals is that these kinds of perceptual experience are as subject to (implicit and perhaps idiosyncratic) beliefs, as are perceptions of the external world. Beyond this, the context of interoception brings about further shifts in how to think about the relationships between body, mind and brain. One such shift is that generative models of interoceptive signals should be geared towards control or regulation of physiological variables, rather than towards accurate representation of some extra-cranial state-of-affairs [7,12,13]. The two goals remain tightly interwoven, inasmuch as effective regulation depends on deployment of sufficiently elaborated predictive models. This shift recognizes alternative origins to ‘Bayesian brain’ ideas in twentieth century ‘cybernetics’ [14,15] and in doing so, points to a deep connection between life and mind, in which cognitive processes are grounded in fundamental evolutionary imperatives to maintain physiological homeostasis [7,16]. Many other specific implications follow, for example in reframing the functional basis of a variety of disorders of emotion and selfhood. Figure 1.Inference and perception across different modalities. Green arrows represent exteroceptive predictions and prediction errors underlying perception of the external world. Orange arrows represent proprioceptive predictions (and prediction errors) generating action through active inference. Blue arrows represent interoceptive predictions (and prediction errors) underlying emotional processing and autonomic regulation. Integrated experiences of embodied selfhood emerge from the joint hierarchical content of self-related predictions across all these dimensions, including—at hierarchically deep levels—multimodal and amodal predictions. Adapted from Seth [7]. (Online version in colour.) Here, we survey these exciting developments. We first provide a brief introduction to the framework of prediction error minimization in the Bayesian brain, emphasizing its embodied or enactive aspects. These aspects appear prominently in the role of action in reducing prediction error (i.e. active inference) and emphasize the key role of uncertainty or precision in shaping the interplay between prior beliefs and sensory evidence. Precision-weighting of recurrent signalling in cortical hierarchies is closely associated with neuromodulation, providing important clues about developmental origins of conditions like autism; it is also associated with attention, suggesting novel accounts for symptom expression due to aberrant attention to interoceptive signals. Turning to functional neuroanatomy, we outline the functional architecture of interoceptive inference and review recent suggestions that perceptual predictions originate preferentially in agranular cortices [9,17], while acknowledging that direct empirical evidence for interoceptive inference is still to be uncovered. We next address some theoretical issues, relating active interoceptive inference to experiences of emotion and embodied selfhood, highlighting a control-oriented or instrumental perspective on interoceptive inference that calls on cybernetic concepts of predictive regulation, allostatic control and perceptual control theory [7,13,18]. We conclude by exploring the implications of these ideas for a sample of clinical conditions that may reflect false interoceptive inference, either in their aetiology and/or in symptom expression. While disorders in emotional processing and interoceptive experience naturally invite explanations in terms of abnormal interoceptive inference, we also highlight how this perspective can illuminate other conditions and symptoms including autism, fatigue and depression. PREDICTIVE CODING IN THE BAYESIAN BRAIN Current formulations of Helmholtz's notion are now the most popular metaphors for neuronal processing and are usually considered under the Bayesian brain hypothesis as predictive coding [6,19–21]. Predictive coding is a process theory with a biologically plausible back story and a considerable amount of empirical support [21,22]. (See [23] for a review of canonical microcircuits and predictive coding in perception, [17,24] for an application of the same ideas to motor control, and [25] for evidence of feed-forward and feed-back signalling carried by distinct frequency bands.) In these schemes, neuronal representations in higher or deeper levels of neuronal hierarchies generate predictions of representations in lower levels. These descending predictions are compared with lower-level representations to form a prediction error (usually associated with the activity of superficial pyramidal cells). This mismatch or difference signal is passed back up the hierarchy, to update higher representations (usually associated with the activity of deep pyramidal cells). The recurrent exchange of signals between adjacent hierarchical levels resolves prediction error at each and every level, resulting in a hierarchically deep explanation for sensory inputs. In computational terms, the activity of neuronal populations is assumed to encode Bayesian beliefs or probability distributions over states in the world that cause sensations (e.g. my visual sensations are caused by a face—see figures 2 and 3). The simplest encoding corresponds to representing the belief with the

Read the complete text