Nonhuman Primate Reaching with Multichannel Sensorimotor Cortex Electrophysiology

General Description. This dataset consists of:

1. The threshold crossing times of extracellularly and simultaneously recorded spikes, sorted into units (up to five, including a "hash" unit), along with sorted waveform snippets, and,
2. The x,y position of the fingertip of the reaching hand and the x,y position of reaching targets (both sampled at 250 Hz).

The behavioral task was to make self-paced reaches to targets arranged in a grid (e.g. 8x8) without gaps or pre-movement delay intervals. One monkey reached with the right arm (recordings made in the left hemisphere); The other reached with the left arm (right hemisphere). In some sessions recordings were made from both M1 and S1 arrays (192 channels); in most sessions M1 recordings were made alone (96 channels).

Data from two primate subjects are included: 37 sessions from monkey 1 ("Indy", spanning about 10 months) and 10 sessions from monkey 2 ("Loco", spanning about 1 month), for a total of ~ 20,000 reaches and 6,500 reaches from monkeys 1 and 2, respectively.

Possible uses. These data are ideal for training BCI decoders, in particular because they are not segmented into trials. We expect that the dataset will be valuable for researchers who wish to design improved models of sensorimotor cortical spiking or provide an equal footing for comparing different BCI decoders. Other uses could include analyses of the statistics of arm kinematics, spike noise-correlations or signal-correlations, or for exploring the stability or variability of extracellular recording over sessions.

Variable names. Each file contains data in the following format. In the below, n refers to the number of recording channels, u refers to the number of sorted units, and k refers to the number of samples.

• chan_names - n x 1
  • A cell array of channel identifier strings, e.g. "M1 001".
• cursor_pos - k x 2
  • The position of the cursor in Cartesian coordinates (x, y), mm.
• finger_pos - k x 3 or k x 6
  • The position of the working fingertip in Cartesian coordinates (z, -x, -y), as reported by the hand tracker in cm. Thus the cursor position is an affine transformation of fingertip position using the following matrix:
    \(\begin{pmatrix} 0 & 0 \\ -10 & 0 \\ 0 & -10 \end{pmatrix}\)
    Note that for some sessions finger_pos includes the orientation of the sensor as well; the full state is thus: (z, -x, -y, azimuth, elevation, roll).
• target_pos - k x 2
  • The position of the target in Cartesian coordinates (x, y), mm.
• t - k x 1
  • The timestamp corresponding to each sample of the cursor_pos, finger_pos, and target_pos, seconds.
• spikes - n x u
  • A cell array of spike event vectors. Each element in the cell array is a vector of spike event timestamps, in seconds. The first unit (u1) is the "unsorted" unit, meaning it contains the threshold crossings which remained after the spikes on that channel were sorted into other units (u2, u3, etc.) For some sessions spikes were sorted into up to 2 units (i.e. u=3); for others, 4 units (u=5).
• wf - n x u
  • A cell array of spike event waveform "snippets". Each element in the cell array is a matrix of spike event waveforms. Each waveform corresponds to a timestamp in "spikes". Waveform samples are in microvolts.

Decoder Results. These data were used to fit decoder models, as reported in Makin, et al [1]. To aid comparisons to other decoders, we include performance summaries (for each session, decoder, bin-width, etc.) in the file refh_results.csv, containing the following columns:

• session - a session identifier, e.g. "indy_20160407_02"
• monkey - one of, "indy" or "loco"
• num_neurons - total number of features used in the decoder
• num_training_samples - number of samples (at the specified bin-width) used to train the decoder (sequential, from file start)
• num_testing_samples - number of samples used to evaluate the decoder (sequential, until file end)
• kinematic_axis - one of, "posx", "posy", "velx", "vely", "accx" or "accy"
• bin_width - one of, "16", "32", "64" or "128"
• decoder - one of, "regression", "KF_observed", "KF_static", "KF_dynamic", "UKF", "rEFH_static" or "rEFH_dynamic"
• rsq - coefficient of determination, R2
• snr - Signal to noise ratio, SNR := -10 log10(1 - R2)

Videos. For some sessions, we recorded screencasts of the stimulus presentation display using a dedicated hardware video grabber. These screencasts are thus a faithful representation of the stimuli and feedback presented to the monkey and are available for the following sessions:

• indy_20160921_01
• indy_20160930_02
• indy_20160930_05
• indy_20161005_06
• indy_20161006_02
• indy_20161007_02
• indy_20161011_03
• indy_20161013_03
• indy_20161014_04
• indy_20161017_02

Supplements. The raw broadband neural recordings that the spike trains in this dataset were extracted from are available for the following sessions:

• indy_20160622_01: doi:10.5281/zenodo.1488440
• indy_20160624_03: doi:10.5281/zenodo.1486147
• indy_20160627_01: doi:10.5281/zenodo.1484824
• indy_20160630_01: doi:10.5281/zenodo.1473703
• indy_20160915_01: doi:10.5281/zenodo.1467953
• indy_20160916_01: doi:10.5281/zenodo.1467050
• indy_20160921_01: doi:10.5281/zenodo.1451793
• indy_20160927_04: doi:10.5281/zenodo.1433942
• indy_20160927_06: doi:10.5281/zenodo.1432818
• indy_20160930_02: doi:10.5281/zenodo.1421880
• indy_20160930_05: doi:10.5281/zenodo.1421310
• indy_20161005_06: doi:10.5281/zenodo.1419774
• indy_20161006_02: doi:10.5281/zenodo.1419172
• indy_20161007_02: doi:10.5281/zenodo.1413592
• indy_20161011_03: doi:10.5281/zenodo.1412635
• indy_20161013_03: doi:10.5281/zenodo.1412094
• indy_20161014_04: doi:10.5281/zenodo.1411978
• indy_20161017_02: doi:10.5281/zenodo.1411882
• indy_20161024_03: doi:10.5281/zenodo.1411474
• indy_20161025_04: doi:10.5281/zenodo.1410423
• indy_20161026_03: doi:10.5281/zenodo.1321264
• indy_20161027_03: doi:10.5281/zenodo.1321256
• indy_20161206_02: doi:10.5281/zenodo.1303720
• indy_20161207_02: doi:10.5281/zenodo.1302866
• indy_20161212_02: doi:10.5281/zenodo.1302832
• indy_20161220_02: doi:10.5281/zenodo.1301045
• indy_20170123_02: doi:10.5281/zenodo.1167965
• indy_20170124_01: doi:10.5281/zenodo.1163026
• indy_20170127_03: doi:10.5281/zenodo.1161225
• indy_20170131_02: doi:10.5281/zenodo.854733

Contact  Information. We would be delighted to hear from you if you find this dataset valuable, especially if it leads to publication. Corresponding author: J. E. O'Doherty <joeyo@neuroengineer.com>.

Citation.

@misc{ODoherty:2017,  author = {O'{D}oherty, Joseph E. and Cardoso, Mariana M. B. and Makin, Joseph G. and Sabes, Philip N.},  title  = {Nonhuman Primate Reaching with Multichannel Sensorimotor Cortex electrophysiology},  doi    = {10.5281/zenodo.788569},  url    = {https://doi.org/10.5281/zenodo.788569},  month  = may,  year   = {2017} }

Publications making use of this dataset.

1. Makin, J. G., O'Doherty, J. E., Cardoso, M. M. B. & Sabes, P. N. (2018). Superior arm-movement decoding from cortex with a new, unsupervised-learning algorithm. J Neural Eng. 15(2): 026010. doi:10.1088/1741-2552/aa9e95
2. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2018). Spike Rate Estimation Using Bayesian Adaptive Kernel Smoother (BAKS) and Its Application to Brain Machine Interfaces. 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Honolulu, HI, USA, 2018, pp. 2547-2550. doi:10.1109/EMBC.2018.8512830
3. Balasubramanian, M., Ruiz, T., Cook, B., Bhattacharyya, S., Prabhat, Shrivastava, A. & Bouchard K. (2018). Optimizing the Union of Intersections LASSO (UoILASSO) and Vector Autoregressive (UoIVAR) Algorithms for Improved Statistical Estimation at Scale. arXiv Preprint. arXiv:1808.06992
4. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2019). Decoding Hand Kinematics from Local Field Potentials Using Long Short-Term Memory (LSTM) Network. arXiv Preprint. arXiv:1901.00708
5. Clark, D. G., Livezey, J. A., & Bouchard, K. E. (2019). Unsupervised Discovery of Temporal Structure in Noisy Data with Dynamical Components Analysis. arXiv Preprint. arXiv:1905.09944
6. Shaikh, S., So, R., Sibindi, T., Libedinsky, C., & Basu, A. (2019). Towards Intelligent Intra-cortical BMI (i2BMI): Low-power Neuromorphic Decoders that outperform Kalman Filters. bioRxiv Preprint. 772988. doi:10.1101/772988
7. Clark, D. G., Livezey, J. A., & Bouchard, K. E. (2019). Unsupervised Discovery of Temporal Structure in Noisy Data with Dynamical Components Analysis. Advances in Neural Information Processing Systems (NeurIPS) 32.
8. Keshtkaran, M. R., & Pandarinath, C. (2019). Enabling hyperparameter optimization in sequential autoencoders for spiking neural data. Advances in Neural Information Processing Systems (NeurIPS) 32.
9. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2019). End-to-End Hand Kinematic Decoding from LFPs Using Temporal Convolutional Network. 2019 IEEE Biomedical Circuits and Systems Conference (BioCAS), Nara, Japan, pp. 1-4. doi:10.1109/biocas.2019.8919131
10. Bose, S. K., Acharya, J., & Basu, A. (2019). Is my Neural Network Neuromorphic? Taxonomy, Recent Trends and Future Directions in Neuromorphic Engineering. 2019 53rd Asilomar Conference on Signals, Systems, and Computers, Pacific Grove, CA, USA, pp. 1522-1527. doi:10.1109/IEEECONF44664.2019.9048891
11. Sachdeva, P. S., Bhattacharyya, S., & Bouchard, K. E. (2019). Sparse, Predictive, and Interpretable Functional Connectomics with UoILasso, 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Berlin, Germany, pp. 1965-1968. doi:10.1109/EMBC.2019.8856316
12. Shaikh, S., So, R., Sibindi, T., Libedinsky, C., & Basu, A. (2019). Towards Intelligent Intracortical BMI (i2BMI): Low-Power Neuromorphic Decoders That Outperform Kalman Filters. IEEE Transactions on Biomedical Circuits and Systems. 13(6): 1615-1624. doi:10.1109/TBCAS.2019.2944486
13. Sachdeva, P. S, Livezey, J. A, Dougherty, M. E., Gu, B.-M., Berke, J. D, & Bouchard, K. E. (2020). Accurate Inference in Parametric Models Reshapes Neuroscientific Interpretation and Improves Data-driven Discovery. bioRxiv Preprint. 2020.04.10.036244. doi:10.1101/2020.04.10.036244
14. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2020). Inferring entire spiking activity from local field potentials with deep learning. bioRxiv Preprint. 2020.05.02.074104. doi:10.1101/2020.05.02.074104
15. Ahmadi, N., Constandinou, T. G., & Bouganis. C.-S. (2020). Impact of referencing scheme on decoding performance of LFP-based brain-machine interface. bioRxiv Preprint. 2020.05.03.075218 doi:10.1101/2020.05.03.075218
16. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2020). Robust and accurate decoding of hand kinematics from entire spiking activity using deep learning. bioRxiv Preprint. 2020.05.07.083063 doi:10.1101/2020.05.07.083063
17. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2020). Improved Spike-based Brain-Machine Interface Using Bayesian Adaptive Kernel Smoother and Deep Learning. TechRxiv Preprint. doi:10.36227/techrxiv.12383600.v1
18. Balasubramanian, M., Ruiz, T., Cook, B., Prabhat, Bhattacharyya, S., Shrivastava, A. & Bouchard K. (2020). Scaling of Union of Intersections for Inference of Granger Causal Networks from Observational Data. Proceeding of the 34th IEEE International Parallel & Distributed Processing Symposium (IPDPS). New Orleans, LA, USA, pp. 264-273. doi: 10.1109/IPDPS47924.2020.00036
19. Keshtkaran, M. R., Sedler, A. R., Chowdhury, R. H., Tandon, R., Basrai, D., Nguyen, S. L, Sohn, H., Jazayeri, M., Miller, L. E., & Pandarinath, C. (2021). A large-scale neural network training framework for generalized estimation of single-trial population dynamics. bioRxiv Preprint. 2021.01.13.426570. doi:10.1101/2021.01.13.426570
20. Ahmadi, N., Constandinou, T. G., & Bouganis. C.-S. (2021). Impact of referencing scheme on decoding performance of LFP-based brain-machine interface. J Neural Eng. 18(1): 016028. doi:10.1088/1741-2552/abce3c
21. Ahmadi, N., Constandinou, T. G., & Bouganis, C.-S. (2021). Robust and accurate decoding of hand kinematics from entire spiking activity using deep learning. J. Neural Eng. 18(2): 026011. doi:10.1088/1741-2552/abde8a
22. Savolainen, O. W. (2021). The Significance of Neural Inter-Frequency Correlations. Research Square Preprint (v1). doi:10.21203/rs.3.rs-329644/v1
23. Sachdeva, P. S., Livezey, J. A., Dougherty, M. E., Gu, B.-M., Berke, J. D., & Bouchard, K. E. (2021). Improved inference in coupling, encoding, and decoding models and its consequence for neuroscientific interpretation. Journal of Neuroscience Methods. 358: 109195. doi:10.1016/j.jneumeth.2021.109195
24. Sani, O. G., Pesaran, B., & Shanechi., M. M. (2021). Where is all the nonlinearity: flexible nonlinear modeling of behaviorally relevant neural dynamics using recurrent neural networks. bioRxiv Preprint. 2021.09.03.458628. doi:10.1101/2021.09.03.458628
25. Yang, S.-H., Huang, J.-W., Huang, C.-J., Chiu, P.-H., Lai, H.-Y., & Chen, Y.-Y. (2021). Selection of Essential Neural Activity Timesteps for Intracortical Brain–Computer Interface Based on Recurrent Neural Network. Sensors. 21(19): 6372. doi:10.3390/s21196372
26. Schimel, M., Kao, T.-C., Jensen, K.T., & Hennequin, G. (2021). iLQR-VAE : control-based learning of input-driven dynamics with applications to neural data. bioRxiv Preprint. 2021.10.07.463540. doi:10.1101/2021.10.07.463540
27. Li, Y., Qi, Y., Wang, Y., Wang, Y., Xu, K., & Pan, G. (2021). Robust neural decoding by kernel regression with Siamese representation learning. J Neural Eng. 18(5): 056062. doi:10.1088/1741-2552/ac2c4e
28. Pei, F., Ye, J., Zoltowski, D., Wu, A., Chowdhury, R. H., Sohn, H., O'Doherty, J. E., Shenoy, K. V., Kaufman, M. T., Churchland, M., Jazayeri, M., Miller, L. E., Pillow, J., Park, I. M., Dyer, E. L., & Pandarinath, C. (2021). Neural Latents Benchmark '21: Evaluating latent variable models of neural population activity. arXiv Preprint. arXiv:2109.04463
29. Savolainen, O.W. (2021). The significance of neural inter-frequency power correlations. Sci. Rep. 11, 23190. doi:10.1038/s41598-021-02277-0
30. Jensen, K. T., Kao, T.-C., Stone, J. T., & Hennequin, G. (2021). Scalable Bayesian GPFA with automatic relevance determination and discrete noise models. bioRxiv Preprint. 2021.06.03.446788. doi:10.1101/2021.06.03.44678
31. Valencia, D., Mercier, P. P, & Alimohammad, A. (2022). In vivo neural spike detection with adaptive noise estimation. J Neural Eng. 19: 046018. doi:10.1088/1741-2552/ac8077
32. Keshtkaran, M. R., Sedler, A. R., Chowdhury, R. H., Tandon, R., Basrai, D., Nguyen, S. L., Sohn, H., Jazayeri, M., Miller, L. E., & Pandarinath, C. (2022). A large-scale neural network training framework for generalized estimation of single-trial population dynamics. Nat Methods. 19, 1572-1577. doi:10.1038/s41592-022-01675-0
33. Qi, Y., Zhu, X., Xu, K., Ren, F., Jiang, H., Zhu, J., Zhang, J., Pan, G., & Wang, Y. (2022). Dynamic Ensemble Bayesian Filter for Robust Control of a Human Brain-Machine Interface. arXiv Preprint. arXiv:2204.11840
34. Qi, Y., Zhu, X., Xu, K., Ren, F., Jiang, H., Zhu, J., Zhang, J., Pan, G., & Wang, Y. (2022). Dynamic Ensemble Bayesian Filter for Robust Control of a Human Brain-Machine Interface. IEEE Transactions on Biomedical Engineering. 69(12): 3825-3835. doi:10.1109/TBME.2022.3182588
35. Zhu, X., Qi, Y., Pan, G., Wang, Y. (2022). Tracking Functional Changes in Nonstationary Signals with Evolutionary Ensemble Bayesian Model for Robust Neural Decoding. Advances in Neural Information Processing Systems (NeurIPS) 35.
36. Ye, J., Collinger, J. L., Wehbe, L., & Gaunt, R. (2023). Neural Data Transformer 2: Multi-Context Pretraining for Neural Spiking Activity. bioRxiv Preprint. 2023.09.18.558113. doi:10.1101/2023.09.18.558113

History.

• Version 2 - added CSV of results from Makin et al.
• Version 1 - initial release.