The Impact of Neural Noise and Expectations on Visual Hallucinations in the Context of Signal Detection Theory and Bayesian Inference

Abstract

Visual hallucinations are perceptions that occur in the absence of a corresponding physical, external, stimulus. While commonly associated with psychotic disorders, like schizophrenia, visual hallucinations also manifest across the healthy population. This suggests that visual hallucinations are not simply the product of a clinical deficit, but rather, they are likely to derive from normal brain functions. As it stands, signal detection theory (SDT) and Bayesian inference (BI) are two leading psychological models that provide distinct perspectives as to how bottom-up processing (e.g., sensory signal) and top-down processing (e.g., cognitive mechanisms, expectations, and prior knowledge) contribute to perception, and hallucination development. SDT posits that expectations do not influence sensory processing, but instead affect cognitive and decisional mechanisms (i.e., top-down processes) that impact behavioural decision strategies. Alternatively, BI models suggest that expectations interact with sensory processing to affect perception under circumstances of poor sensory precision. Notably, the quality of our sensory processing can be affected by various normal intrinsic mechanisms in the healthy brain. Neural noise is a one neurobiological characteristic that is able to reduce the quality of sensory processing. There is a wealth of research highlighting the co-occurrence of neural noise and visual hallucinations across various contexts and populations. Therefore, it may be a strong candidate mechanism to more specifically explain how visual hallucinations arise in healthy and clinical populations. However, given the limited techniques with a demonstrated capacity to safely increase neural noise in the human brain, there is little evidence to suggest a direct causal role in the development of hallucinations. Although, transcranial random noise stimulation (tRNS) is an emerging, non-invasive, electrical brain stimulation technique that may allow us modulate neural noise in humans. The current thesis therefore aims to extend on research investigating the mechanistic effects of tRNS to explore whether it can be used to safely increase neural noise in the healthy brain. Subsequently, we aimed to apply tRNS to more directly explore the causal role of neural noise in the development of visual hallucinations in the context expectations under SDT and BI theoretical models. Show more