Research Overview

The Embodied Computation Group investigates how visceral and embodied processes shape our decision-making, emotion, and conscious perception. Our research program combines computational modelling, psychophysics, physiological measures, and advanced neuroimaging.

Methods

Our research uses an integrated set of methods spanning computation, brain imaging, and physiology:

Visceral Inference & Interoception

Visceral Inference & Interoception

How does the brain process signals from within the body, and how do our beliefs and expectations shape our embodied self experience? We study the computational and neural mechanisms underlying interoception - the sense of the internal state of the body. Our work examines how cardiac, respiratory, and gastric signals influence perception, attention, and decision-making. Using computational and psychophysical techniques, as well as maker methods like 3D printing, we develop open source, robustly validated tools for quantifying interoception across different domains. This enables us to understand how beliefs and expectations are integrated with visceral signals in different contexts and populations.

Representative Publications:

Computational Models of Brain-Body Interaction

Computational Models of Brain-Body Interaction

We develop and test computational models that formalize how the brain integrates bodily signals with external sensory information. Drawing on predictive processing and active inference frameworks, we aim to understand the computational principles governing embodied cognition.

Representative Publications:

Emotion & Affect

Emotion & Affect

Emotions are fundamentally embodied experiences. We investigate how bodily states contribute to emotional experience and how disruptions in brain-body interaction may contribute to affective disorders such as anxiety and depression.

Representative Publications:

Consciousness, Perception, and Metacognition

Consciousness, Perception, and Metacognition

How do we form confidence in our decisions and perceptions? Our work spans perceptual and value-based decision making to understand how we form confidence representations and the computational principles underlying metacognition. We develop new computational models of metacognition across cognitive, perceptual, and interoceptive domains. A key focus is understanding how these processes interact with brain-body dynamics and interoception, including how cardiac and respiratory cycles modulate conscious perception and awareness.

Representative Publications:

Causal Interventions in Interoception

Causal Interventions in Interoception

Developing causal manipulations of neurovisceral connections and visceral afferents is essential to understand and disentangle the intricate web of brain-body connections that govern emotion, consciousness, and mental health. In this work, we develop pharmacological, behavioral, and lifestyle interventions, as well as emerging neural stimulation techniques including vagal nerve stimulation and ultrasound stimulation, to causally perturb specific neural and visceral foci and understand their role in interoception and beyond.

Representative Publications:

Clinical Applications & Computational Psychiatry

Clinical Applications & Computational Psychiatry

We bring together our models, manipulations, and measures to map, understand, phenotype, and ultimately intervene upon disorders of mood, affect, and brain-body interaction. This work is particularly focused on interoception as a bridge between somatic and psychiatric illnesses and their strong comorbidities. For example, patients with psychiatric illnesses are more than twice as likely to die from cancer and cardiometabolic disease, and suffer substantially worse outcomes and delays in their journey through the healthcare system. We believe interoception plays a key role in this and are actively applying our research to help patients from both sides achieve better outcomes.

Representative Publications:

Visit our publications page for a complete list of our research outputs.