Building and rebuilding models of self and world: From sensory plasticity to psychosis

How does the brain construct a stable sense of self and world when sensory experience is radically altered? A growing body of work shows that perception is not a passive readout of sensory input, but an active process in which the brain builds and continuously updates models of the body and environment. In this talk, I will present a research programme investigating how these models are shaped by sensory plasticity, challenged by technological augmentation, and potentially disrupted in psychiatric conditions.
I will first draw on my work in congenital blindness to show that the absence of vision leads to selective and systematic changes in bodily and perceptual processing. Blind individuals exhibit enhanced access to interoceptive signals, altered affective touch perception, and reduced susceptibility to multisensory bodily illusions, alongside structural reorganisation in occipital cortex linked to interoceptive accuracy. These findings suggest that sensory deprivation does not only enhance remaining senses, but recalibrates how information about the body is weighted and integrated.
I will then turn to my recent work on wearable finger-extension devices as a model of rapid plasticity. These studies demonstrate that artificial body extensions can be integrated into proprioceptive and tactile representations within minutes, following structured and dynamic changes in body maps. Together, these findings reveal that models of the body are not fixed representations, but flexible and continuously updated constructs shaped by both long-term experience and short-term sensorimotor interactions.
Finally, I will outline ongoing work examining whether such plasticity may confer resilience or vulnerability in the domain of mental health. Congenital blindness has been proposed to protect against psychosis, yet the mechanisms remain unclear. I will present a framework grounded in predictive processing, proposing that blindness recalibrates the balance between prior expectations and sensory evidence, yielding more stable inferential models, whereas psychosis may reflect a disruption of this balance. By combining computational modelling and neuroimaging across blind, sighted, and clinical populations, this work aims to identify computational and neural markers of adaptive and maladaptive perceptual inference.
Together, this work shows that sensory experience shapes how the brain calibrates and stabilises models of self and world, with implications for technology integration and psychiatric resilience.

Join at: imt.lu/sagrestia