C'est visiblement mal étudié encore, mais ça semble équivalent au tabac ou pire. Respirer des produits de combustion n'est jamais bon... Voici un exemple de méta-analyse en français:
doi.org/10.1016/j.rmra.2024.11.128

4/4
➡️ Read our manuscript on the link below 🔽🔽🔽
➡️ Many thanks to our collaborators at L2S @centralesupelec.bsky.social , and to all members of the @touchmovelab.bsky.social‬ for their key input to this project.
➡️ Funding from Fondation 3DS and @agencerecherche.bsky.social

3/4 As learning progressed, the mice learned to coordinate the prosthesis movements across dimensions, and they managed to consolidate a dominant prosthesis trajectory that led to rewards.

2/4 Our prosthesis can perform movements in a 3D space next to the body of the mice, that control it via a chronic invasive brain-machine interface. We show that the mice can drive it in a 2D subspace, while some mice managed to control it also in a 3D space.

6/6 This is all thanks to the incredibly talented and passionate Anton Dogadov, who carried this whole study!

Thanks for feedback on our first version of the paper from @gbouvier.bsky.social vier.bsky.social‬ at @neuropsi.bsky.social, Jean-François Léger @ibensens.bsky.social and many others!

5/6 During cortical wave generation, the response to the movement was reshaped, and lost its relation to limb amplitude, likely being processed to match the task instead.

4/6 Importantly, limb movements were likely required to trigger rewardable waves, but were only successful when the cortex was in the specific, suppressed state.

3/6 We have now gone much further, and show that learning this task triggers a reshaping of the cortical dynamics, including a strong suppression of activity before rewardable wave.

2/6 One year ago, we showed that mice can learn to guide mesoscale cortical of the somatosensory cortex from a start to a target area to obtain rewards.