Decoding semantic representations from functional near-infrared spectroscopy signals

• DOI: 10.1117/1.nph.5.1.011003
• Publication Date: 2017-08-24T01:22:14Z
• Authors: Benjamin D. Zinszer; Laurie Bayet; Lauren L. Emberson; Rajeev D. S. Raizada; Richard N. Aslin
• Keywords: Similarity (geometry); Functional near-infrared spectroscopy; Representation

This study uses representational similarity-based neural decoding to test whether semantic information elicited by words and pictures is encoded in functional near-infrared spectroscopy (fNIRS) data. In experiment 1, subjects passively viewed eight audiovisual word picture stimuli for 15 min. Blood oxygen levels were measured using the Hitachi ETG-4000 fNIRS system with a posterior array over occipital lobe left lateral temporal lobe. Each participant's response patterns abstracted similarity space compared group average (excluding that subject, i.e., leave-one-out cross-validation) distributional model of representation. Mean accuracy both tasks significantly exceeded chance. 2, we three group-level models averaging structures from sets participants each group. these models, was accurately decoded model, while between-groups comparison. Our findings indicate representations are data, preserved across subjects, decodable an extrinsic model. These results first attempt link pattern higher-level how related other.