What happens in our brains when we have to make quick decisions based on uncertain signals? The Bayesian Brain Framework (BBF) proposes that the human brain functions as an inferential system constantly engaged in making predictions about its surroundings based on received sensory input. In this context, predictive coding describes perception as an active process of comparing sensory input with top-down predictions based on internal probabilistic models (Friston, 2010). These predictions are in constant formation and are modified whenever the comparison produces a discrepancy. Although there are numerous behavioral studies supporting this theory, the neurophysiological correlates underlying these mechanisms are a subject for further study. This thesis aims to investigate such neural bases through the use of transcranial random noise electrical stimulation (tRNS) during the performance of the Two-Flash Fusion (TFF) task. We hypothesize that modulation by tRNS of neural activity in the alpha band in the occipital areas (V1/V2 and V5/MT) may influence the ability of the perceptual system to sample sensory input. In addition, we hypothesize that stimulation of the theta band in medial prefrontal areas (mPFC) may influence the ability to extract statistical regularities to create the generative models and to use these predictive models. The results showed little influence of tRNS on the processes investigated in both occipital and frontal areas. This demonstrates a greater complexity of the processes involved in predictive coding than hypothesized. This finding provides information regarding the modulability of the mechanisms and suggests the need to adopt a focal-spectrum stimulation technique, suitable for alpha and theta bands.
Cosa accade nel nostro cervello quando dobbiamo prendere decisioni rapide basate su segnali incerti? Il Bayesian Brain Framework (BBF) propone che il cervello umano funzioni come un sistema inferenziale costantemente impegnato a formulare predizioni sull’ambiente circostante in base agli input sensoriali ricevuti. In questo contesto, la codifica predittiva descrive la percezione come un processo attivo di confronto tra input sensoriali e predizioni top-down, basate su modelli probabilistici interni (Friston, 2010). Queste predizioni sono in costante formazione e vengono modificate ogni volta che il confronto produce una discrepanza. Sebbene esistano numerosi studi comportamentali a supporto di questa teoria, i correlati neurofisiologici che sottendono tali meccanismi sono oggetto di studio da approfondire. Questa tesi si propone di indagare tali basi neurali attraverso l’utilizzo della stimolazione elettrica transcranica random noise (tRNS) durante l’esecuzione del compito di Two-Flash Fusion (TFF). Si ipotizza che la modulazione tramite tRNS dell’attività neurale nella banda alpha nelle aree occipitale (V1/V2 e V5/MT) possa influenzare la capacità del sistema percettivo di campionare gli input sensoriali. Inoltre, si ipotizza che la stimolazione della banda theta nelle aree prefrontali mediali (mPFC) possa influenzare la capacità di estrarre le regolarità statistiche per creare i modelli generativi e di utilizzare tali modelli predittivi. I risultati hanno mostrato una scarsa influenza della tRNS sui processi indagati, sia nelle aree occipitali che frontali. Ciò dimostra una maggiore complessità dei processi coinvolti nella codifica predittiva rispetto a quelli ipotizzati. Tale conclusione fornisce informazioni riguardo la modulabilità dei meccanismi e suggerisce la necessità di adottare una tecnica di stimolazione a spettro focale, adatta alle bande alpha e theta.
Predire per percepire: Analisi dell'influenza della tRNS sulla codifica predittiva nella visione
ROUBAL, LETIZIA
2024/2025
Abstract
What happens in our brains when we have to make quick decisions based on uncertain signals? The Bayesian Brain Framework (BBF) proposes that the human brain functions as an inferential system constantly engaged in making predictions about its surroundings based on received sensory input. In this context, predictive coding describes perception as an active process of comparing sensory input with top-down predictions based on internal probabilistic models (Friston, 2010). These predictions are in constant formation and are modified whenever the comparison produces a discrepancy. Although there are numerous behavioral studies supporting this theory, the neurophysiological correlates underlying these mechanisms are a subject for further study. This thesis aims to investigate such neural bases through the use of transcranial random noise electrical stimulation (tRNS) during the performance of the Two-Flash Fusion (TFF) task. We hypothesize that modulation by tRNS of neural activity in the alpha band in the occipital areas (V1/V2 and V5/MT) may influence the ability of the perceptual system to sample sensory input. In addition, we hypothesize that stimulation of the theta band in medial prefrontal areas (mPFC) may influence the ability to extract statistical regularities to create the generative models and to use these predictive models. The results showed little influence of tRNS on the processes investigated in both occipital and frontal areas. This demonstrates a greater complexity of the processes involved in predictive coding than hypothesized. This finding provides information regarding the modulability of the mechanisms and suggests the need to adopt a focal-spectrum stimulation technique, suitable for alpha and theta bands.| File | Dimensione | Formato | |
|---|---|---|---|
|
Letizia_Roubal.pdf
accesso aperto
Dimensione
787.04 kB
Formato
Adobe PDF
|
787.04 kB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/86753