Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus
When reading a sentence, individual words can be combined to create more complex meaning. In this study, we sought to uncover brain regions that reflect the representation of the meaning of sentences at the topic level, as opposed to the meaning of their individual constituent words when considered...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-05-01
|
| Series: | NeuroImage |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1053811922001343 |
| _version_ | 1818338850065350656 |
|---|---|
| author | David J. Acunzo Daniel M. Low Scott L. Fairhall |
| author_facet | David J. Acunzo Daniel M. Low Scott L. Fairhall |
| author_sort | David J. Acunzo |
| collection | DOAJ |
| description | When reading a sentence, individual words can be combined to create more complex meaning. In this study, we sought to uncover brain regions that reflect the representation of the meaning of sentences at the topic level, as opposed to the meaning of their individual constituent words when considered irrespective of their context. Using fMRI, we recorded the neural activity of participants while reading sentences. We constructed a topic-level sentence representations using the final layer of a convolutional neural network (CNN) trained to classify Wikipedia sentences into broad semantic categories. This model was contrasted with word-level sentence representations constructed using the average of the word embeddings constituting the sentence. Using representational similarity analysis, we found that the medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus more strongly represent sentence topic-level, compared to word-level, meaning, uncovering the important role of these semantic system regions in the representation of topic-level meaning. Results were comparable when sentence meaning was modelled with a multilayer perceptron that was not sensitive to word order within a sentence, suggesting that the learning objective, in the terms of the topic being modelled, is the critical factor in capturing these neural representational spaces. |
| first_indexed | 2024-12-13T15:17:39Z |
| format | Article |
| id | doaj.art-91be5555c2b84abfb8aeeb45c3ff282f |
| institution | Directory Open Access Journal |
| issn | 1095-9572 |
| language | English |
| last_indexed | 2024-12-13T15:17:39Z |
| publishDate | 2022-05-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage |
| spelling | doaj.art-91be5555c2b84abfb8aeeb45c3ff282f2022-12-21T23:40:39ZengElsevierNeuroImage1095-95722022-05-01251119005Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrusDavid J. Acunzo0Daniel M. Low1Scott L. Fairhall2CIMeC/University of Trento, Corso Bettini 31, Rovereto 38068, ItalyProgram in Speech and Hearing Bioscience and Technology, Harvard Medical School, United States; Brain and Cognitive Sciences Department, MIT, United StatesCIMeC/University of Trento, Corso Bettini 31, Rovereto 38068, Italy; Corresponding author.When reading a sentence, individual words can be combined to create more complex meaning. In this study, we sought to uncover brain regions that reflect the representation of the meaning of sentences at the topic level, as opposed to the meaning of their individual constituent words when considered irrespective of their context. Using fMRI, we recorded the neural activity of participants while reading sentences. We constructed a topic-level sentence representations using the final layer of a convolutional neural network (CNN) trained to classify Wikipedia sentences into broad semantic categories. This model was contrasted with word-level sentence representations constructed using the average of the word embeddings constituting the sentence. Using representational similarity analysis, we found that the medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus more strongly represent sentence topic-level, compared to word-level, meaning, uncovering the important role of these semantic system regions in the representation of topic-level meaning. Results were comparable when sentence meaning was modelled with a multilayer perceptron that was not sensitive to word order within a sentence, suggesting that the learning objective, in the terms of the topic being modelled, is the critical factor in capturing these neural representational spaces.http://www.sciencedirect.com/science/article/pii/S1053811922001343Representational similarity analysisConvolutional neural networkfMRISentence processingSemantic representationSemantic system |
| spellingShingle | David J. Acunzo Daniel M. Low Scott L. Fairhall Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus NeuroImage Representational similarity analysis Convolutional neural network fMRI Sentence processing Semantic representation Semantic system |
| title | Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus |
| title_full | Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus |
| title_fullStr | Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus |
| title_full_unstemmed | Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus |
| title_short | Deep neural networks reveal topic-level representations of sentences in medial prefrontal cortex, lateral anterior temporal lobe, precuneus, and angular gyrus |
| title_sort | deep neural networks reveal topic level representations of sentences in medial prefrontal cortex lateral anterior temporal lobe precuneus and angular gyrus |
| topic | Representational similarity analysis Convolutional neural network fMRI Sentence processing Semantic representation Semantic system |
| url | http://www.sciencedirect.com/science/article/pii/S1053811922001343 |
| work_keys_str_mv | AT davidjacunzo deepneuralnetworksrevealtopiclevelrepresentationsofsentencesinmedialprefrontalcortexlateralanteriortemporallobeprecuneusandangulargyrus AT danielmlow deepneuralnetworksrevealtopiclevelrepresentationsofsentencesinmedialprefrontalcortexlateralanteriortemporallobeprecuneusandangulargyrus AT scottlfairhall deepneuralnetworksrevealtopiclevelrepresentationsofsentencesinmedialprefrontalcortexlateralanteriortemporallobeprecuneusandangulargyrus |