Inter-individual deep image reconstruction via hierarchical neural code conversion

• Published in: NeuroImage (Orlando, Fla.) Vol. 271; p. 120007
• Main Authors: Ho, Jun Kai, Horikawa, Tomoyasu, Majima, Kei, Cheng, Fan, Kamitani, Yukiyasu
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.05.2023; Elsevier Limited; Elsevier
• Subjects: Activity patterns; Adult; Brain; Brain Mapping; Correspondence; Decoding; Deep Learning; fMRI; Functional alignment; Functional magnetic resonance imaging; Humans; Image processing; Image Processing, Computer-Assisted - methods; Machine learning; Magnetic Resonance Imaging; Male; Neural networks; Neuroimaging; Sensorimotor Cortex - anatomy & histology; Somatosensory cortex; Topography; Visual cortex; Visual hierarchy; Visual image reconstruction; Young Adult; Functional alignment; fMRI; Decoding; Visual hierarchy; Visual image reconstruction

•Neural code converters, which are trained to predict brain activity patterns from one to another individual when presented with the same stimulus, automatically learn the hierarchical correspondence of visual areas.•Converted brain activity patterns can be decoded into hierarchical DNN features to...