An emerging view of neural geometry in motor cortex supports high-performance decoding
Decoders for brain-computer interfaces (BCIs) assume constraints on neural activity, chosen to reflect scientific beliefs while yielding tractable computations. Recent scientific advances suggest that the true constraints on neural activity, especially its geometry, may be quite different from those...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
eLife Sciences Publications Ltd
2025-02-01
|
| Series: | eLife |
| Subjects: | |
| Online Access: | https://elifesciences.org/articles/89421 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832542821023744000 |
|---|---|
| author | Sean M Perkins Elom A Amematsro John Cunningham Qi Wang Mark M Churchland |
| author_facet | Sean M Perkins Elom A Amematsro John Cunningham Qi Wang Mark M Churchland |
| author_sort | Sean M Perkins |
| collection | DOAJ |
| description | Decoders for brain-computer interfaces (BCIs) assume constraints on neural activity, chosen to reflect scientific beliefs while yielding tractable computations. Recent scientific advances suggest that the true constraints on neural activity, especially its geometry, may be quite different from those assumed by most decoders. We designed a decoder, MINT, to embrace statistical constraints that are potentially more appropriate. If those constraints are accurate, MINT should outperform standard methods that explicitly make different assumptions. Additionally, MINT should be competitive with expressive machine learning methods that can implicitly learn constraints from data. MINT performed well across tasks, suggesting its assumptions are well-matched to the data. MINT outperformed other interpretable methods in every comparison we made. MINT outperformed expressive machine learning methods in 37 of 42 comparisons. MINT’s computations are simple, scale favorably with increasing neuron counts, and yield interpretable quantities such as data likelihoods. MINT’s performance and simplicity suggest it may be a strong candidate for many BCI applications. |
| format | Article |
| id | doaj-art-589e05e126a54471a379c4a1dcb525c7 |
| institution | Kabale University |
| issn | 2050-084X |
| language | English |
| publishDate | 2025-02-01 |
| publisher | eLife Sciences Publications Ltd |
| record_format | Article |
| series | eLife |
| spelling | doaj-art-589e05e126a54471a379c4a1dcb525c72025-02-03T14:30:15ZengeLife Sciences Publications LtdeLife2050-084X2025-02-011210.7554/eLife.89421An emerging view of neural geometry in motor cortex supports high-performance decodingSean M Perkins0https://orcid.org/0000-0001-9456-4648Elom A Amematsro1https://orcid.org/0000-0003-4843-4513John Cunningham2Qi Wang3Mark M Churchland4https://orcid.org/0000-0001-9123-6526Department of Biomedical Engineering, Columbia University, New York, United States; Zuckerman Institute, Columbia University, New York, United StatesZuckerman Institute, Columbia University, New York, United States; Department of Neuroscience, Columbia University Medical Center, New York, United StatesZuckerman Institute, Columbia University, New York, United States; Department of Statistics, Columbia University, New York, United States; Center for Theoretical Neuroscience, Columbia University Medical Center, New York, United States; Grossman Center for the Statistics of Mind, Columbia University, New York, United StatesDepartment of Biomedical Engineering, Columbia University, New York, United StatesZuckerman Institute, Columbia University, New York, United States; Department of Neuroscience, Columbia University Medical Center, New York, United States; Grossman Center for the Statistics of Mind, Columbia University, New York, United States; Kavli Institute for Brain Science, Columbia University Medical Center, New York, United StatesDecoders for brain-computer interfaces (BCIs) assume constraints on neural activity, chosen to reflect scientific beliefs while yielding tractable computations. Recent scientific advances suggest that the true constraints on neural activity, especially its geometry, may be quite different from those assumed by most decoders. We designed a decoder, MINT, to embrace statistical constraints that are potentially more appropriate. If those constraints are accurate, MINT should outperform standard methods that explicitly make different assumptions. Additionally, MINT should be competitive with expressive machine learning methods that can implicitly learn constraints from data. MINT performed well across tasks, suggesting its assumptions are well-matched to the data. MINT outperformed other interpretable methods in every comparison we made. MINT outperformed expressive machine learning methods in 37 of 42 comparisons. MINT’s computations are simple, scale favorably with increasing neuron counts, and yield interpretable quantities such as data likelihoods. MINT’s performance and simplicity suggest it may be a strong candidate for many BCI applications.https://elifesciences.org/articles/89421brain-computer interfacemotor controldecodingpopulation activitystate estimation |
| spellingShingle | Sean M Perkins Elom A Amematsro John Cunningham Qi Wang Mark M Churchland An emerging view of neural geometry in motor cortex supports high-performance decoding eLife brain-computer interface motor control decoding population activity state estimation |
| title | An emerging view of neural geometry in motor cortex supports high-performance decoding |
| title_full | An emerging view of neural geometry in motor cortex supports high-performance decoding |
| title_fullStr | An emerging view of neural geometry in motor cortex supports high-performance decoding |
| title_full_unstemmed | An emerging view of neural geometry in motor cortex supports high-performance decoding |
| title_short | An emerging view of neural geometry in motor cortex supports high-performance decoding |
| title_sort | emerging view of neural geometry in motor cortex supports high performance decoding |
| topic | brain-computer interface motor control decoding population activity state estimation |
| url | https://elifesciences.org/articles/89421 |
| work_keys_str_mv | AT seanmperkins anemergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT elomaamematsro anemergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT johncunningham anemergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT qiwang anemergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT markmchurchland anemergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT seanmperkins emergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT elomaamematsro emergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT johncunningham emergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT qiwang emergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding AT markmchurchland emergingviewofneuralgeometryinmotorcortexsupportshighperformancedecoding |