1
|
Tamura H. Acquisition of similar properties by filters in the same stream of a multistream convolutional neural network. Sci Rep 2025; 15:67. [PMID: 39747548 PMCID: PMC11697301 DOI: 10.1038/s41598-024-84981-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/16/2024] [Accepted: 12/30/2024] [Indexed: 01/04/2025] Open
Abstract
Functional modular organization is observed in a variety of cortical areas in the brain. In the visual cortex of primates, adjacent neurons often respond to the same visual submodality, such as color or orientation, and have a similar preferred orientation or preferred color. However, it remains unclear why functional modular organization emerges in the cerebral cortex. In the present study, I constructed and trained a multistream convolutional neural network to examine whether filters in the same stream acquire similar properties. Although filters in the same stream were able to develop any structures, they acquired similar degrees of orientation and color selectivity and preferred similar orientations and colors. The deletion of filters in a single stream that had similar degrees of stimulus selectivity resulted in larger decreases in classification accuracy than the deletion of those that did not. By contrast, the deletion of filters in a single stream that shared a preferred stimulus resulted in similar decreases in classification accuracy to the deletion of those that did not. Together, these findings suggest that filters with similar degrees of stimulus selectivity in the same stream are required for optimal task performance of the multistream convolutional neural network, and probably of the brain.
Collapse
Affiliation(s)
- Hiroshi Tamura
- Graduate School of Frontier Biosciences, The University of Osaka, Suita, Osaka, 565-0871, Japan.
- Center for Information and Neural Networks, Suita, Osaka, 565-0871, Japan.
- Cognitive Neuroscience Group, Graduate School of Frontier Biosciences, The University of Osaka, 1-4 Yamadaoka, Suita, Osaka, 565-0871, Japan.
| |
Collapse
|
2
|
Koizumi K, Inoue M, Chowdhury S, Bito H, Yamanaka A, Ishizuka T, Yawo H. Functional emergence of a column-like architecture in layer 5 of mouse somatosensory cortex in vivo. J Physiol Sci 2019; 69:65-77. [PMID: 29761270 PMCID: PMC10716991 DOI: 10.1007/s12576-018-0618-4] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/14/2017] [Accepted: 04/30/2018] [Indexed: 01/04/2023]
Abstract
To investigate how the functional architecture is organized in layer 5 (L5) of the somatosensory cortex of a mouse in vivo, the input-output relationship was investigated using an all-optical approach. The neural activity in L5 was optically recorded using a Ca2+ sensor, R-CaMP2, through a microprism inserted in the cortex under two-photon microscopy, while the L5 was regionally excited using optogenetics. The excitability was spread around the blue-light irradiated region, but the horizontal propagation was limited to within a certain distance (λ < 130 μm from the center of the illumination spot). When two regions were photostimulated with a short interval, the excitability of each cluster was reduced. Therefore, a column-like architecture had functionally emerged with reciprocal inhibition through a minimal number of synaptic relays. This could generate a synchronous output from a region of L5 with simultaneous enhancement of the signal-to-noise ratio by silencing of the neighboring regions.
Collapse
Affiliation(s)
- Kyo Koizumi
- Department of Developmental Biology and Neuroscience, Tohoku University Graduate School of Life Sciences, 2-1-1 Katahira, Aoba-ku, Sendai, 980-8577, Japan
| | - Masatoshi Inoue
- Department of Neurochemistry, Graduate School of Medicine, University of Tokyo, Tokyo, Japan
- Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Tokyo, Japan
| | - Srikanta Chowdhury
- Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Tokyo, Japan
- Department of Neuroscience II, Research Institute of Environmental Medicine, Nagoya University, Nagoya, Japan
| | - Haruhiko Bito
- Department of Neurochemistry, Graduate School of Medicine, University of Tokyo, Tokyo, Japan
- Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Tokyo, Japan
| | - Akihiro Yamanaka
- Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Tokyo, Japan
- Department of Neuroscience II, Research Institute of Environmental Medicine, Nagoya University, Nagoya, Japan
| | - Toru Ishizuka
- Department of Developmental Biology and Neuroscience, Tohoku University Graduate School of Life Sciences, 2-1-1 Katahira, Aoba-ku, Sendai, 980-8577, Japan
| | - Hiromu Yawo
- Department of Developmental Biology and Neuroscience, Tohoku University Graduate School of Life Sciences, 2-1-1 Katahira, Aoba-ku, Sendai, 980-8577, Japan.
- Center for Neuroscience, Tohoku University Graduate School of Medicine, Sendai, Japan.
| |
Collapse
|