Weakly supervised learning in neural encoding for the position of the moving finger of a macaque

  • PDF / 2,084,173 Bytes
  • 14 Pages / 595.276 x 790.866 pts Page_size
  • 104 Downloads / 144 Views

DOWNLOAD

REPORT


Weakly supervised learning in neural encoding for the position of the moving finger of a macaque Jingyi Feng 1 & Haifeng Wu 1

&

Yu Zeng 1 & Yuhong Wang 1

Received: 23 July 2019 / Accepted: 10 June 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract The problem of neural decoding is essential for the realization of a neural interface. In this study, the position of the moving finger of a macaque was directly decoded through the neuron spike signals in the motor cortex, instead of relying on the synergy of the related muscle tissues around the body, also known as neural decoding. Currently, supervised learning is the most commonly employed method for this purpose. However, based on existing technologies, unsupervised learning with regression causes excessive errors. To solve this problem, weakly supervised learning (WSL) was used to correct the predicted position of the moving finger of a macaque in unsupervised training. Then, the corrected finger position was further used to train and accurately fit the weight parameters. We then utilized public data to evaluate the decoding performance of the Kalman filter (KF) and the expectation maximization (EM) algorithms in the WSL model. Unlike in previous methods, in WSL, the only available information is that the finger has moved to four areas in the plane, instead of the actual track value. When compared to the supervised models, the WSL decoding performance only differs by approximately 0.4%. This result improves by 41.3% relative to unsupervised models in the two-dimensional plane. The investigated approach overcomes the instability and inaccuracy of unsupervised learning. What’s more, the method in the paper also verified that the unsupervised encoding and decoding technology of neuronal signals is related to the range of external activities, rather than having a priori specific location. Keywords Neural decoding . Macaque moving finger . Unsupervised learning . Weakly supervised learning . Kalman filter . Expectation maximization

Introduction At present, the realization of brain-like intelligence is essential for the study of the nervous system for the behavior and perception of the external world. Prosthetics, robots, and other devices that fully realize “brain control technology” through psychological means are becoming a reality [1]. For example, tactile mice and wearable technologies can be given to * Haifeng Wu [email protected] Jingyi Feng [email protected] Yu Zeng [email protected] Yuhong Wang [email protected] 1

School of Electrical and Information Technology, Yunnan Minzu University, City One, Kunming, China

individuals who suffer from comprehensive impairment of the visual and auditory channels. This will enable such individuals to exchange information with the external environment [2]. Additionally, electroencephalogram and electromyography signals have been used to study the nonlinear in the brain to achieve a real-time manual reconstruction system for humeral amputees [3]. These technologies not only have dire