Artificial neural network modelling of the neural population code underlying mathematical operations
Mathematical operations have long been regarded as a sparse, symbolic process in neuroimaging studies. In contrast, advances in artificial neural networks (ANN) have enabled extracting distributed representations of mathematical operations. Recent neuroimaging studies have compared distributed repre...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-04-01
|
| Series: | NeuroImage |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S105381192300126X |
| _version_ | 1797869860551655424 |
|---|---|
| author | Tomoya Nakai Shinji Nishimoto |
| author_facet | Tomoya Nakai Shinji Nishimoto |
| author_sort | Tomoya Nakai |
| collection | DOAJ |
| description | Mathematical operations have long been regarded as a sparse, symbolic process in neuroimaging studies. In contrast, advances in artificial neural networks (ANN) have enabled extracting distributed representations of mathematical operations. Recent neuroimaging studies have compared distributed representations of the visual, auditory and language domains in ANNs and biological neural networks (BNNs). However, such a relationship has not yet been examined in mathematics. Here we hypothesise that ANN-based distributed representations can explain brain activity patterns of symbolic mathematical operations. We used the fMRI data of a series of mathematical problems with nine different combinations of operators to construct voxel-wise encoding/decoding models using both sparse operator and latent ANN features. Representational similarity analysis demonstrated shared representations between ANN and BNN, an effect particularly evident in the intraparietal sulcus. Feature-brain similarity (FBS) analysis served to reconstruct a sparse representation of mathematical operations based on distributed ANN features in each cortical voxel. Such reconstruction was more efficient when using features from deeper ANN layers. Moreover, latent ANN features allowed the decoding of novel operators not used during model training from brain activity. The current study provides novel insights into the neural code underlying mathematical thought. |
| first_indexed | 2024-04-10T00:18:10Z |
| format | Article |
| id | doaj.art-b878db9895904b22a3626ee44afb3b2c |
| institution | Directory Open Access Journal |
| issn | 1095-9572 |
| language | English |
| last_indexed | 2024-04-10T00:18:10Z |
| publishDate | 2023-04-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage |
| spelling | doaj.art-b878db9895904b22a3626ee44afb3b2c2023-03-16T05:03:09ZengElsevierNeuroImage1095-95722023-04-01270119980Artificial neural network modelling of the neural population code underlying mathematical operationsTomoya Nakai0Shinji Nishimoto1Center for Information and Neural Networks, National Institute of Information and Communications Technology, Suita, Japan; Lyon Neuroscience Research Center (CRNL), INSERM U1028 - CNRS UMR5292, University of Lyon, Bron, France; Corresponding author at: Centre Hospitalier Le Vinatier (Bât. 452), 95 Bd Pinel, 69500 Bron, France.Center for Information and Neural Networks, National Institute of Information and Communications Technology, Suita, Japan; Graduate School of Frontier Biosciences, Osaka University, Suita, Japan; Graduate School of Medicine, Osaka University, Suita, JapanMathematical operations have long been regarded as a sparse, symbolic process in neuroimaging studies. In contrast, advances in artificial neural networks (ANN) have enabled extracting distributed representations of mathematical operations. Recent neuroimaging studies have compared distributed representations of the visual, auditory and language domains in ANNs and biological neural networks (BNNs). However, such a relationship has not yet been examined in mathematics. Here we hypothesise that ANN-based distributed representations can explain brain activity patterns of symbolic mathematical operations. We used the fMRI data of a series of mathematical problems with nine different combinations of operators to construct voxel-wise encoding/decoding models using both sparse operator and latent ANN features. Representational similarity analysis demonstrated shared representations between ANN and BNN, an effect particularly evident in the intraparietal sulcus. Feature-brain similarity (FBS) analysis served to reconstruct a sparse representation of mathematical operations based on distributed ANN features in each cortical voxel. Such reconstruction was more efficient when using features from deeper ANN layers. Moreover, latent ANN features allowed the decoding of novel operators not used during model training from brain activity. The current study provides novel insights into the neural code underlying mathematical thought.http://www.sciencedirect.com/science/article/pii/S105381192300126XfMRIMathematicsIPSEncoding modelArtificial neural networkDecoding model |
| spellingShingle | Tomoya Nakai Shinji Nishimoto Artificial neural network modelling of the neural population code underlying mathematical operations NeuroImage fMRI Mathematics IPS Encoding model Artificial neural network Decoding model |
| title | Artificial neural network modelling of the neural population code underlying mathematical operations |
| title_full | Artificial neural network modelling of the neural population code underlying mathematical operations |
| title_fullStr | Artificial neural network modelling of the neural population code underlying mathematical operations |
| title_full_unstemmed | Artificial neural network modelling of the neural population code underlying mathematical operations |
| title_short | Artificial neural network modelling of the neural population code underlying mathematical operations |
| title_sort | artificial neural network modelling of the neural population code underlying mathematical operations |
| topic | fMRI Mathematics IPS Encoding model Artificial neural network Decoding model |
| url | http://www.sciencedirect.com/science/article/pii/S105381192300126X |
| work_keys_str_mv | AT tomoyanakai artificialneuralnetworkmodellingoftheneuralpopulationcodeunderlyingmathematicaloperations AT shinjinishimoto artificialneuralnetworkmodellingoftheneuralpopulationcodeunderlyingmathematicaloperations |