Clinical translation of a high-performance neural prosthesis

Gilja, Vikash; Pandarinath, Chethan; Blabe, Christine H; Nuyujukian, Paul; Simeral, John D.; Sarma, Anish A.; Sorice, Brittany L; Perge, János A.; Jarosiewicz, Beata; Hochberg, Leigh R.; Shenoy, Krishna V.; Henderson, Jaimie M.

doi:10.1038/nm.3953

Cited by 295 publications

(304 citation statements)

References 32 publications

Supporting

Mentioning

295

Contrasting

Unclassified

Order By: Relevance

“…The copyright holder for this preprint .training the LFADS model parameters is a computationally demanding task, running the LFADS model involves a series of linear transformations and static nonlinearities, which are on the same order of computational complexity as existing BMI implementations that run in real-time (Sussillo et al 2012;Gilja et al 2015;) . Second, the LFADS architecture must be modified to be causal, i.e., the current neural state estimate must depend only on previously observed data.…”

Section: Discussionmentioning

confidence: 99%

Inferring single-trial neural population dynamics using sequential auto-encoders

Pandarinath

O’Shea

Collins

et al. 2017

Preprint

Self Cite

196

411

View full text Add to dashboard Cite

Neuroscience is experiencing a data revolution in which simultaneous recording of many hundreds or thousands of neurons is revealing structure in population activity that is not apparent from single-neuron responses. This structure is typically extracted from trial-averaged data. Single-trial analyses are challenging due to incomplete sampling of the neural population, trial-to-trial variability, and fluctuations in action potential timing. Here we introduce Latent Factor Analysis via Dynamical Systems (LFADS), a deep learning method to infer latent dynamics from single-trial neural spiking data. LFADS uses a nonlinear dynamical system (a recurrent neural network) to infer the dynamics underlying observed population activity and to extract 'de-noised' single-trial firing rates from neural spiking data. We apply LFADS to a variety of monkey and human motor cortical datasets, demonstrating its ability to predict observed behavioral variables with unprecedented accuracy, extract precise estimates of neural dynamics on single trials, infer perturbations to those dynamics that correlate with behavioral choices, and combine data from non-overlapping recording sessions (spanning months) to improve inference of underlying dynamics. In summary, LFADS leverages all observations of a neural population's activity to accurately model its dynamics on single trials, opening the door to a detailed understanding of the role of dynamics in performing computation and ultimately driving behavior.Increasing evidence suggests that in many brain areas, such as the motor and prefrontal cortices, the activity of large populations of neurons, termed the neural population state, is often well-described by low-dimensional dynamics [e.g. (Afshar et al. 2011; Harvey, Coen, and Tank 2012; Kaufman et al. 2014;Sadtler et al. 2014;Kobak et al. 2016a) ]. Recovering these dynamics on single trials is essential for illuminating the relationship between neural population activity and behavior, and for advancing therapeutic neurotechnologies such as closed-loop deep brain stimulation and brain-machine interfaces. However, recovering population dynamics All rights reserved. No reuse allowed without permission.(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/152884 doi: bioRxiv preprint first posted online Jun. 20, 2017; on single trials is difficult due to trial-to-trial variability (e.g. in behavior or arousal) and fluctuations in the spiking of individual neurons. Even with dramatic increases in the numbers of neurons that can be simultaneously recorded using multichannel electrode arrays or optical imaging, accurately recovering population dynamics from single trials remains a significant challenge for data-analysis methods.Standard analyses sacrifice single-trial information for the sake of better estimates of trial-averaged neural states (Ahrens et al. 2012; Kobak et al. 2016b) . Techniques for extrac...

show abstract

Section: Discussionmentioning

confidence: 99%

Inferring single-trial neural population dynamics using sequential auto-encoders

Pandarinath

O’Shea

Collins

et al. 2017

Preprint

Self Cite

196

411

View full text Add to dashboard Cite

show abstract

“…For example, tetraplegic patients have been able to volitionally control robotic arms [29,49], while other patients have been able to control their own limbs via iBCI-based approaches [37,91]. In addition to restoration of motor movement, enhancing communication, as in the case of ALS patients, for example, can also be achieved by iBCI [15,42,55].…”

Section: Brain Computer Interfacesmentioning

confidence: 99%

Interfacing with the nervous system: a review of current bioelectric technologies

et al. 2017

View full text Add to dashboard Cite

The aim of this study is to discuss the state of the art with regard to established or promising bioelectric therapies meant to alter or control neurologic function. We present recent reports on bioelectric technologies that interface with the nervous system at three potential sites-(1) the end organ, (2) the peripheral nervous system, and (3) the central nervous system-while exploring practical and clinical considerations.

show abstract

“…1 and 2E) that are free of problem-causing percutaneous leads and connectors [17]. Wireless operation is key for experimental neuroscience with freely behaving subjects [4], as well as potential translation to human use [27]. However, we have also been exploring high-speed serial communication over FDAcleared implantable leads and connectors that are commonly used in cardiac pacing and neuromodulation applications [28] (Figs.…”

Section: B Communication and Power Interfacesmentioning

confidence: 99%

Integrated neural interfaces

Walker

Rieth

Iyer

et al. 2017

2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS)

View full text Add to dashboard Cite

Abstract-Electronic interfaces to the nervous system are increasingly important for experimental neuroscience as well as medical diagnostics and therapies. Existing neural interfaces are limited, however, by the use of passive recording and stimulation devices that are connected to active electronics on separate physical platforms through large amounts of passive wiring. This manuscript proposes a new approach to integrate active electronics directly with neural recording and stimulation devices using state-of-the-art silicon processing, assembly, and packaging techniques. System concepts are described for fully wireless operation as well as wireline interfacing through FDA-cleared implantable leads and connectors. The approach offers a more modular paradigm of neural interface design, which is greatly needed as the demand for higher channel counts grows.

show abstract

Clinical translation of a high-performance neural prosthesis

Cited by 295 publications

References 32 publications

Inferring single-trial neural population dynamics using sequential auto-encoders

Inferring single-trial neural population dynamics using sequential auto-encoders

Interfacing with the nervous system: a review of current bioelectric technologies

Integrated neural interfaces

Contact Info

Product

Resources

About