CaltechAUTHORS
  A Caltech Library Service

An Active Learning Algorithm for Control of Epidural Electrostimulation

Desautels, Thomas A. and Choe, Jaehoon and Gad, Parag and Nandra, Mandheerej S. and Roy, Roland R. and Zhong, Hui and Tai, Yu-Chong and Edgerton, V. Reggie and Burdick, Joel W. (2015) An Active Learning Algorithm for Control of Epidural Electrostimulation. IEEE Transactions on Biomedical Engineering, 62 (10). pp. 2443-2455. ISSN 0018-9294. PMCID PMC4617183. http://resolver.caltech.edu/CaltechAUTHORS:20150922-132655726

[img] PDF - Accepted Version
See Usage Policy.

1091Kb

Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechAUTHORS:20150922-132655726

Abstract

Epidural electrostimulation has shown promise for spinal cord injury therapy. However, finding effective stimuli on the multi-electrode stimulating arrays employed requires a laborious manual search of a vast space for each patient. Widespread clinical application of these techniques would be greatly facilitated by an autonomous, algorithmic system which choses stimuli to simultaneously deliver effective therapy and explore this space. We propose a method based on GP-BUCB, a Gaussian process bandit algorithm. In n = 4 spinally transected rats, we implant epidural electrode arrays and examine the algorithm’s performance in selecting bipolar stimuli to elicit specified muscle responses. These responses are compared with temporally interleaved intra-animal stimulus selections by a human expert. GP-BUCB successfully controlled the spinal electrostimulation preparation in 37 testing sessions, selecting 670 stimuli. These sessions included sustained autonomous operations (ten-session duration). Delivered performance with respect to the specified metric was as good as or better than that of the human expert. Despite receiving no information as to anatomically likely locations of effective stimuli, GP-BUCB also consistently discovered such a pattern. Further, GP-BUCB was able to extrapolate from previous sessions’ results to make predictions about performance in new testing sessions, while remaining sufficiently flexible to capture temporal variability. These results provide validation for applying automated stimulus selection methods to the problem of spinal cord injury therapy.


Item Type:Article
Related URLs:
URLURL TypeDescription
http://dx.doi.org/10.1109/TBME.2015.2431911 DOIArticle
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7105837PublisherArticle
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4617183PubMed CentralArticle
ORCID:
AuthorORCID
Tai, Yu-Chong0000-0001-8529-106X
Additional Information:© 2015 IEEE. Manuscript received December 15, 2014; revised March 30, 2015; accepted April 28, 2015. Date of publication May 12, 2015; date of current version September 16, 2015. This work was supported by under Grant NIH U01EB15521, Grant R01EB007615, the Leona M. and Harry B. Helmsley Charitable Trust, the Christopher and Dana Reeve, Broccoli, Walkabout, and F. M. Kirby Foundations. J. Choe and P. Gad contributed equally to this work. The authors would like to thank Y. Sui and M. Rath for their assistance in executing the experiments. V. R. Edgerton, R. R. Roy, and J. W. Burdick hold shareholder interest in NeuroRecovery Technologies (NRT) and hold certain inventorship rights on intellectual property licensed by The Regents of the University of California to NRT and its subsidiaries. V. R. Edgerton is also the President and Chairman of the Board. A patent has been submitted covering concepts described here.
Funders:
Funding AgencyGrant Number
NIHU01EB15521
NIHR01EB007615
Leona M. and Harry B. Helmsley Charitable TrustUNSPECIFIED
Christopher and Dana Reeve FoundationUNSPECIFIED
Broccoli FoundationUNSPECIFIED
Walkabout FoundationUNSPECIFIED
F. M. Kirby FoundationUNSPECIFIED
Subject Keywords:Implants, learning automata, neural engineering, neuromuscular stimulation, spinal cord injury (SCI)
PubMed Central ID:PMC4617183
Record Number:CaltechAUTHORS:20150922-132655726
Persistent URL:http://resolver.caltech.edu/CaltechAUTHORS:20150922-132655726
Official Citation:Desautels, T.A.; Choe, J.; Gad, P.; Nandra, M.S.; Roy, R.R.; Zhong, H.; Tai, Y.; Edgerton, V.R.; Burdick, J.W., "An Active Learning Algorithm for Control of Epidural Electrostimulation," in Biomedical Engineering, IEEE Transactions on , vol.62, no.10, pp.2443-2455, Oct. 2015 doi: 10.1109/TBME.2015.2431911
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:60420
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:22 Sep 2015 22:32
Last Modified:19 Jul 2017 21:10

Repository Staff Only: item control page