Recent advantages of brain decoding with functional magnetic resonance imaging (fMRI) have enabled us to estimate individual differences in mental information from brain responses to natural sensory inputs. However, the physical constraints and costs of fMRI measurements prevent brain decoding from achieving real-world applications. To address this issue, this study aims to build a framework to decode individual differences in mental information under natural situations via brain-response prediction using convolutional neural networks (CNNs). Once the CNN-based prediction model is constructed using measured brain response, no additional fMRI measurements are needed to decode mental information from the predicted responses of individual brains. As per our analysis, it was found that in 81 of 87 items to be decoded, this framework captured individual difference patterns consistent with conventional decoding using measured brain responses. Our framework has great potential to decode personal mental information with minimal fMRI measuring constraints or costs, which substantially expands the applicability of brain decoding in daily life.
