Back

Presence Hallucination Induction through Robotically Mediated Somatomotor Conflicts: a pooled analysis of 26 experiments

Dhanis, H.; Potheegadoo, J.; Faivre, N.; Hara, M.; Bernasconi, F.; Blanke, O.

2025-09-28 neuroscience
10.1101/2025.09.26.678742 bioRxiv
Show abstract

Hallucinations are significant symptoms in psychiatric and neurodegenerative diseases, that may indicate advanced disease progression or worse disease forms. They are also frequent in healthy individuals, especially elderly or bereaved. Despite their relevance, studying hallucinations in controlled laboratory conditions remains challenging given the limited procedures inducing clinically relevant hallucinations. We have previously developed a robotics-based protocol capable of inducing a specific clinically relevant hallucination, presence hallucination (PH), in both healthy individuals and patients. Using this approach, we have systematically investigated the sensitivity and intensity of PH-induction, as well as its effects on various sensory, behavioral and cognitive aspects. In the present study, we pooled individual-participant data from 26 in-house experiments (totaling 580 individuals) and conducted a Bayesian analysis to estimate effects and moderators of PH-induction. PH-induction was reliably induced with a medium effect-size, and individuals with schizotypal traits were more sensitive. Furthermore, we identified that PH-induction may not strongly depend on altered agency, but found a synergistic relationship between passivity and induced PH. Collectively these results elucidate the role of somatomotor processes in aberrant own body perceptions, advance understanding of psychosis, and provide powerful statistical priors for future studies.

Published in Cortex (predicted rank #1) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.