Neural network-based encoding in free-viewing fMRI with precision models

Representations learned by convolutional neural networks (CNNs) exhibit a remarkable resemblance to information processing patterns observed in the primate visual system on large neuroimaging datasets collected under diverse, naturalistic visual stimulation but with instruction for participants to maintain central fixation. However, this condition diverges significantly from ecologically valid visual behaviour, suppresses activity in visually active regions, and imposes substantial cognitive load on the viewing task. We present a modification of the encoding model framework, adapting it for use with naturalistic vision datasets acquired under fully natural viewing conditions – without fixation – by incorporating eye-tracking data and receptive field maps. Our precision encoding models were trained on the StudyForrest dataset, which features fully naturalistic movie viewing. By combining voxel-specific population receptive field estimates with eye-tracking data for each frame, we generate subjectand voxel-specific feature time series. These time series are constructed by sampling only the locally and temporally relevant elements of the CNN feature map for each voxel. Our results demonstrate that precision encoding models outperform conventional encoding models. This framework provides a strong foundation and justification for future large-scale data collection under fixation-free, fully naturalistic viewing conditions.