Neural Encoding of Densely Sampled fMRI Voxel Responses to Naturalistic Audio-Visual Stimuli
| dc.contributor.advisor | Güçlü, U. | |
| dc.contributor.advisor | Gerven, Marcel, van | |
| dc.contributor.author | Sommers, Rowan | |
| dc.date.issued | 2017-08-01 | |
| dc.description.abstract | Visual object recognition is achieved by the brain’s visual ventral stream. For the purpose of mapping the increasingly complex receptive fields along the hierarchically organized ventral pathway, new DNN analysis techniques can be used. Voxel receptive fields are measured by first predicting voxel BOLD responses from DNN layer features, and subsequently analysing the layers with the highest prediction accuracy for each voxel. The current study tested whether the prediction accuracies of previous models can be improved upon by applying this analysis technique on a newly collected, densely sampled dataset containing fMRI responses to naturalistic audio-visual stimuli. Our results show that the two-step encoding model employed here was able to predict the voxel BOLD responses reasonably well. With small improvements to the analysis methodology, the collected dataset is expected to lead to state-of-the-art prediction accuracies and new insights into the receptive fields of the visual ventral stream. | en_US |
| dc.embargo.lift | 2043-08-01 | |
| dc.embargo.type | Tijdelijk embargo | en_US |
| dc.identifier.uri | https://theses.ubn.ru.nl/handle/123456789/7732 | |
| dc.language.iso | en | en_US |
| dc.thesis.faculty | Faculteit der Sociale Wetenschappen | en_US |
| dc.thesis.specialisation | Researchmaster Cognitive Neuroscience | en_US |
| dc.thesis.studyprogramme | Researchmaster Cognitive Neuroscience | en_US |
| dc.thesis.type | Researchmaster | en_US |
| dc.title | Neural Encoding of Densely Sampled fMRI Voxel Responses to Naturalistic Audio-Visual Stimuli | en_US |