Activations of deep convolutional neural networks are aligned with gamma band activity of human visual cortex

• Published in: Communications biology Vol. 1; no. 1; p. 107
• Main Authors: Kuzovkin, Ilya, Vicente, Raul, Petton, Mathilde, Lachaux, Jean-Philippe, Baciu, Monica, Kahane, Philippe, Rheims, Sylvain, Vidal, Juan R., Aru, Jaan
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 01.01.2018; Nature Publishing Group
• Subjects: 631/378/116/2396; 631/378/2613/2616; 9/30; Artificial intelligence; Biology; Biomedical and Life Sciences; Cognitive science; Information processing; Life Sciences; Neural networks; Neuroscience; Pattern recognition; Psychology; Visual cortex
• ISSN: 2399-3642; 2399-3642
• DOI: 10.1038/s42003-018-0110-y

Recent advances in the field of artificial intelligence have revealed principles about neural processing, in particular about vision. Previous work demonstrated a direct correspondence between the hierarchy of the human visual areas and layers of deep convolutional neural networks (DCNN) trained on visual object recognition. We use DCNN to investigate which frequency bands correlate with feature transformations of increasing complexity along the ventral visual pathway. By capitalizing on intracranial depth recordings from 100 patients we assess the alignment between the DCNN and signals at different frequency bands. We find that gamma activity (30–70 Hz) matches the increasing complexity of visual feature representations in DCNN. These findings show that the activity of the DCNN captures the essential characteristics of biological object recognition not only in space and time, but also in the frequency domain. These results demonstrate the potential that artificial intelligence algorithms have in advancing our understanding of the brain. Ilya Kuzovkin et al. compare intracranial depth recordings from human subjects taken during a visual recognition task to activations of deep convolutional neural networks (DCNNs). They find that signals in gamma frequency bands in the recordings are aligned with the hierarchical layer structure of the DCNN, showing that DCNNs capture important characteristics of biological object recognition.