Probability Cloud Regression Reveals Hidden Variables in Psychedelic Neuroimaging (PsiConnect Pre-Registration)

Abstract

This study investigates whether measurement error in neuroimaging data contains meaningful structure rather than being purely noise. Specifically, it evaluates a novel statistical framework—Probability Cloud Regression (PCR)—which models each observation as a probability distribution rather than a fixed point, allowing uncertainty to be incorporated directly into analysis. Using the publicly available PsiConnect dataset (multimodal fMRI and EEG under psilocybin vs baseline conditions), the project tests whether PCR can: Recover stronger brain–behaviour relationships than traditional regression methods Reveal structured variance patterns consistent with theoretical models of psychedelic brain function (e.g., the REBUS model and entropic brain hypothesis) Identify latent neural regimes and error structures that differ across conditions and subgroups The core innovation of this work lies in reframing measurement error as informative signal about underlying system dynamics, rather than a nuisance to be removed. Expected outcomes include: Demonstration of attenuation bias correction in neuroimaging analyses Evidence that psychedelic states alter not just neural signals, but the structure of measurement variance itself Identification of latent neural regimes through unsupervised methods applied to denoised data If successful, this work would introduce a new methodological paradigm for analysing noisy scientific data and provide empirical support for integrating uncertainty-aware models into neuroscience and beyond. 2. Purpose of the Study The purpose of this study is to evaluate whether modelling measurement uncertainty explicitly improves inference in complex biological systems. Specifically, it aims to determine whether Probability Cloud Regression can extract meaningful structure from noisy multimodal neuroimaging data and provide insights into the neural dynamics of altered states of consciousness. 3. Expected Outcomes It is expected that: PCR will produce less biased and stronger brain–behaviour relationships than OLS Psychedelic conditions will exhibit more complex and structured variance patterns Latent neural regimes will be more recoverable after PCR-based denoising More broadly, the study expects to show that measurement error contains structured information that can be systematically analysed.

Topics & conditions

Detected automatically by Psychedelic Commons from the title and abstract, not labels from the source. They may be incomplete.

Keywords