From Prosthesis to proprioceptive Perception via neuromechanical modelling and Optogenetic cortical stimulation

01/02/2026 to 31/01/2029

Accurate limb movement and embodiment are vital bi-directional brain-body interactions that depend on
proprioceptive feedback. Conditions where proprioception is severed include amputation, diabetes and
several other diseases that lead to peripheral sensory neuropathy (Li et al., 2019; Makin & Flor, 2020). The
loss of proprioceptive inputs leads to severe mobility impairments (e.g. ataxia, preliminary data in Fig. 1A-C)
and has a profound impact on brain physiology linked to maladaptive reorganizations of somatosensory
representations at the cortical level (Makin & Flor, 2020). In particular, it has been linked to a loss of limb
embodiment and build-up of perceptual illusions (Schone et al., 2022), including phantom limb pain, which
manifests in up to 80% of amputees (Giummarra et al., 2008). In contrast to tactile feedback (Flesher et al.,
2021) which already reduces phantom limb pain (Petrini et al., 2019), there are currently no strategies to
address proprioceptive loss and restore the underlying bi-directional brain-body interaction. This is partly
due to the scattered representations of proprioceptive inputs at the surface of the cortex (Alonso et al.,
2023), which prevents the targeted activation of specific proprioceptive percepts with the current clinical
approach based on mesoscale electrical cortical stimulation.
Here we aim to explore alternative proprioceptive feedback strategies that would take advantage of cortical
plasticity. Our hypothesis is that cortical feedback that does not match the detailed encoding of the targeted
cortical area can still provide proprioceptive information to the subject if it abides to the general cortical
encoding constraints that we uncovered in our earlier studies regarding cortical stimulation (Abbasi et al.,
2023; Bagur et al., 2025). Our overall goal with this project is to build a new framework to encode
proprioceptive feedback through biomimetic models of limb mechanics and neural processing, and to test
in mice if this approach yields efficient restoration of proprioception and embodiment of a prosthesis.