Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes

Yu Chieh Lin, Chin Chou, Shin Hung Yang, Hsin Yi Lai, Yu Chun Lo, You Yin Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Changes in the functional mapping between neural activities and kinematic parameters over time poses a challenge to current neural decoder of brain machine interfaces (BMIs). Traditional decoders robust to changes in functional mappings required many day's training data. The decoder may not be robust when it was trained by data from only few days. Therefore, a decoder should be trained to handle a variety of neural-to-kinematic mappings using limited training data. We proposed an evolutionary neural network with error feedback, ECPNN-EF, as a neural decoder, that considered the previous error as an input to the decoder in order to improve the robustness. The decoder was validated to reconstruct rat's forelimb movement in a water-reward lever-pressing task. Two days of data were only used to train the decoder while ten days of data were used to test the decoder. The results showed that the performance of ECPNN-EF was significantly higher than that of standard recurrent neural network without error feedback, which was commonly used in BMI. This suggested that ECPNN-EF trained with few days of training data can be robust to changes in functional mappings.

Original languageEnglish
Title of host publication40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2539-2542
Number of pages4
Volume2018-July
ISBN (Electronic)9781538636466
DOIs
Publication statusPublished - Oct 26 2018
Event40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 - Honolulu, United States
Duration: Jul 18 2018Jul 21 2018

Conference

Conference40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018
CountryUnited States
CityHonolulu
Period7/18/187/21/18

Fingerprint

Brain-Computer Interfaces
Forelimb
Biomechanical Phenomena
Decoding
Trajectories
Neural networks
Feedback
Reward
Brain
Kinematics
Water
Recurrent neural networks
Rats
Formative Feedback

ASJC Scopus subject areas

  • Signal Processing
  • Biomedical Engineering
  • Computer Vision and Pattern Recognition
  • Health Informatics

Cite this

Lin, Y. C., Chou, C., Yang, S. H., Lai, H. Y., Lo, Y. C., & Chen, Y. Y. (2018). Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes. In 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 (Vol. 2018-July, pp. 2539-2542). [8512775] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EMBC.2018.8512775

Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes. / Lin, Yu Chieh; Chou, Chin; Yang, Shin Hung; Lai, Hsin Yi; Lo, Yu Chun; Chen, You Yin.

40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. Vol. 2018-July Institute of Electrical and Electronics Engineers Inc., 2018. p. 2539-2542 8512775.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lin, YC, Chou, C, Yang, SH, Lai, HY, Lo, YC & Chen, YY 2018, Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes. in 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. vol. 2018-July, 8512775, Institute of Electrical and Electronics Engineers Inc., pp. 2539-2542, 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018, Honolulu, United States, 7/18/18. https://doi.org/10.1109/EMBC.2018.8512775
Lin YC, Chou C, Yang SH, Lai HY, Lo YC, Chen YY. Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes. In 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. Vol. 2018-July. Institute of Electrical and Electronics Engineers Inc. 2018. p. 2539-2542. 8512775 https://doi.org/10.1109/EMBC.2018.8512775
Lin, Yu Chieh ; Chou, Chin ; Yang, Shin Hung ; Lai, Hsin Yi ; Lo, Yu Chun ; Chen, You Yin. / Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes. 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. Vol. 2018-July Institute of Electrical and Electronics Engineers Inc., 2018. pp. 2539-2542
@inproceedings{d2f04b91958c4d229be56d9854f8f2dc,
title = "Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes",
abstract = "Changes in the functional mapping between neural activities and kinematic parameters over time poses a challenge to current neural decoder of brain machine interfaces (BMIs). Traditional decoders robust to changes in functional mappings required many day's training data. The decoder may not be robust when it was trained by data from only few days. Therefore, a decoder should be trained to handle a variety of neural-to-kinematic mappings using limited training data. We proposed an evolutionary neural network with error feedback, ECPNN-EF, as a neural decoder, that considered the previous error as an input to the decoder in order to improve the robustness. The decoder was validated to reconstruct rat's forelimb movement in a water-reward lever-pressing task. Two days of data were only used to train the decoder while ten days of data were used to test the decoder. The results showed that the performance of ECPNN-EF was significantly higher than that of standard recurrent neural network without error feedback, which was commonly used in BMI. This suggested that ECPNN-EF trained with few days of training data can be robust to changes in functional mappings.",
author = "Lin, {Yu Chieh} and Chin Chou and Yang, {Shin Hung} and Lai, {Hsin Yi} and Lo, {Yu Chun} and Chen, {You Yin}",
year = "2018",
month = "10",
day = "26",
doi = "10.1109/EMBC.2018.8512775",
language = "English",
volume = "2018-July",
pages = "2539--2542",
booktitle = "40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Neural Decoding Forelimb Trajectory Using Evolutionary Neural Networks with Feedback-Error-Learning Schemes

AU - Lin, Yu Chieh

AU - Chou, Chin

AU - Yang, Shin Hung

AU - Lai, Hsin Yi

AU - Lo, Yu Chun

AU - Chen, You Yin

PY - 2018/10/26

Y1 - 2018/10/26

N2 - Changes in the functional mapping between neural activities and kinematic parameters over time poses a challenge to current neural decoder of brain machine interfaces (BMIs). Traditional decoders robust to changes in functional mappings required many day's training data. The decoder may not be robust when it was trained by data from only few days. Therefore, a decoder should be trained to handle a variety of neural-to-kinematic mappings using limited training data. We proposed an evolutionary neural network with error feedback, ECPNN-EF, as a neural decoder, that considered the previous error as an input to the decoder in order to improve the robustness. The decoder was validated to reconstruct rat's forelimb movement in a water-reward lever-pressing task. Two days of data were only used to train the decoder while ten days of data were used to test the decoder. The results showed that the performance of ECPNN-EF was significantly higher than that of standard recurrent neural network without error feedback, which was commonly used in BMI. This suggested that ECPNN-EF trained with few days of training data can be robust to changes in functional mappings.

AB - Changes in the functional mapping between neural activities and kinematic parameters over time poses a challenge to current neural decoder of brain machine interfaces (BMIs). Traditional decoders robust to changes in functional mappings required many day's training data. The decoder may not be robust when it was trained by data from only few days. Therefore, a decoder should be trained to handle a variety of neural-to-kinematic mappings using limited training data. We proposed an evolutionary neural network with error feedback, ECPNN-EF, as a neural decoder, that considered the previous error as an input to the decoder in order to improve the robustness. The decoder was validated to reconstruct rat's forelimb movement in a water-reward lever-pressing task. Two days of data were only used to train the decoder while ten days of data were used to test the decoder. The results showed that the performance of ECPNN-EF was significantly higher than that of standard recurrent neural network without error feedback, which was commonly used in BMI. This suggested that ECPNN-EF trained with few days of training data can be robust to changes in functional mappings.

UR - http://www.scopus.com/inward/record.url?scp=85056638253&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056638253&partnerID=8YFLogxK

U2 - 10.1109/EMBC.2018.8512775

DO - 10.1109/EMBC.2018.8512775

M3 - Conference contribution

C2 - 30440925

AN - SCOPUS:85056638253

VL - 2018-July

SP - 2539

EP - 2542

BT - 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -