concept

posterior region of interest

Also known as: posterior region of interest, posterior regions of interest

Facts (10)

Sources

Adversarial testing of global neuronal workspace and ... - Nature nature.com Nature Apr 30, 2025 10 facts

measurementIn an adversarial study, researchers compared the performance of decoders trained exclusively on posterior regions of interest (ROIs) with those trained on posterior and prefrontal ROIs together, finding that adding prefrontal ROIs did not improve, and in some cases reduced, category and orientation decoding accuracy.

measurementDecoding of face orientation was robust from MEG cortical time series in posterior regions of interest (approximately 75% with pseudotrial aggregation), but was weaker (35%) in prefrontal regions of interest.

measurementThe iEEG analysis utilized 576 electrodes for the prefrontal region of interest (ROI) and 583 electrodes for the posterior region of interest (ROI).

measurementIn posterior regions of interest, face–object decoding demonstrated significant cross-task generalization with more than 95% accuracy for the approximate duration of the stimulus.

measurementAmong 31 patients, 29.5% (194 of 657) of posterior region of interest (ROI) electrodes and 18.7% (123 of 655) of prefrontal cortex (PFC) ROI electrodes exhibited high gamma responses to stimuli.

claimIn the study 'Adversarial testing of global neuronal workspace and integrated information theory', cross-task generalization of face–object decoding was significant in both posterior and prefrontal regions of interest (ROIs) within theory-predicted time windows.

measurementDecoding of face orientation (left, right, or front views) was achieved in posterior but not in prefrontal regions of interest using iEEG (approximately 95% with pseudotrial aggregation) and fMRI searchlight approach (approximately 45%).

claimAcross all modalities, orientation decoding was observed for letters and false fonts—but not objects—in posterior, but not prefrontal, regions of interest.

claimAcross modalities, face–object decoding occurred in both posterior and prefrontal regions of interest, which is consistent with predictions from both Integrated Information Theory (IIT) and Global Neuronal Workspace Theory (GNWT).

measurementThe study utilized a sample of 29 iEEG patients for decoding analyses, with 576 electrodes in PFC regions of interest and 583 electrodes in posterior regions of interest, and further analyzed a population of 65 healthy participants using MEG.