Hideaki Hirose

Prediction of arm trajectory from a small number of neuron activities in the primary motor cortex

Monkey arm movement was reconstructed from neuron activities recorded in the primary motor cortex (Ml). We recorded single neuron activities from a monkeys Ml, while the animal performed an arm reaching task. We also recorded electromyographic (EMG) activity and movement trajectories during the task. First, we reconstructed the EMG signals from the neuron activities. The EMG signals were reliably reconstructed with a linear summation of the neuron activities. Next, we reconstructed joint angles from the reconstructed EMG signals with an artificial neural network model. The reconstructed trajectories of the hand position and elbow position showed good correlation with the actual arm movement. This model appropriately reflected the anatomical characteristics.

2009 Special Issue: Prediction of arm trajectory from the neural activities of the primary motor cortex with modular connectionist architecture

In our previous study [Koike, Y., Hirose, H., Sakurai, Y., Iijima T., (2006). Prediction of arm trajectory from a small number of neuron activities in the primary motor cortex. Neuroscience Research, 55, 146-153], we succeeded in reconstructing muscle activities from the offline combination of single neuron activities recorded in a serial manner in the primary motor cortex of a monkey and in reconstructing the joint angles from the reconstructed muscle activities during a movement condition using an artificial neural network. However, the joint angles during a static condition were not reconstructed. The difficulties of reconstruction under both static and movement conditions mainly arise due to muscle properties such as the velocity-tension relationship and the length-tension relationship. In this study, in order to overcome the limitations due to these muscle properties, we divided an artificial neural network into two networks: one for movement control and the other for posture control. We also trained the gating network to switch between the two neural networks. As a result, the gating network switched the modules properly, and the accuracy of the estimated angles improved compared to the case of using only one artificial neural network.

Prediction of four degrees of freedom arm movement using EMG signal

In this paper, we predicted four degrees of freedom movement of a monkeys arm by using a neural network model based on the EMG signal. Through the monkeys reaching task, we measured the electromyograms (EMG) signal from the seven muscles of the arm and simultaneously three dimensional movement trajectory of it. The neural network model used in this study is composed of three layers: the input layer with seven values from the EMG signal of the seven muscles, the middle layer consisted of ten and the output layer of four outputs. The movement predicted by this model was almost the same as the real movement. Besides we could implement no delay interface using the EMG signal that is a fundamental signal from the brain that makes it possible to induce the bodys movement. Moreover we can predict not only the external movement of the monkeys arm but also the force of it, which is impossible to be sensed by the external movement sensing devices

Prediction of Arm Trajectory from the Neural Activities of the Primary Motor Cortex Using a Modular Artificial Neural Network Model

First, we reconstructed 9 muscle tensions (filtered EMG signals) from 105 neurons in the arm region of the primary motor cortex, then estimated arm movement in four degrees of freedom in the shoulder and the elbow from the reconstructed 9 muscle tensions. The reconstructed arm movement showed good correlation with the actual arm movement.

Cortical activation during reaching movements in the monkey as revealed by optical recording

To examine the process that prepares for forthcoming movement associated with object vision in the dorsal premotor cortex (dorsal PM), we investigated neuronal activities in the dorsal PM of two rhesus monkeys that performed a delayed-isometric force exertion task. In this task, the subject exerted one of two magnitudes of force by isometric wrist flexion after a 3-s preparatory period; each force condition was instructed by one of two different cue objects prior to the preparatory period. Of 112 neurons with preparatory-period activity, 41 showed a visual preference; i.e., the magnitude of the preparatoryperiod activity varied significantly with the different visual cues associated with the same force condition. Furthermore, half of these (n = 19/41,46%) showed both a visual preference and a significant, independent correlation(s) with one or more movement-related parameter(s). This group of neurons may be involved in motor-preparation processes associated with object vision. Thus, the dorsal PM appears to play a role in linking object vision with the process of preparing for movements associated with the visual object.

Real-time optical monitoring of brain activity of behaving monkey

By increasing the static magnetic field from 1ST to 4T or higher, the signal generally increases because I) the current induced in the sensing coil by sinusoidal magnetic signal is proportional to the signal frequency. which is in turn proportional to the magnitude of static field, and 2) the polarization of protons is proportional to the static field. This increase of signal improves the signal to noise ratio. In addition, because the BOLD effect from veins and venules is proportional to the static field whereas that from capillaries is proportional to the square of the static field up to around 4T. the signal source becomes more localized at higher fields. Based on these two factors, BOLD fMR1 is expected to have higher spatial resolution at higher static fields. fMR1 at higher static fields also has disadvantages. T2* of protons is shorter at higher fields, and thus the time to measure the signal after a single rf pulse is more limited. Single shot EPI is thus more difficult. The susceptibility artifacts due to the inhomogeneity of static field around the nasal cavity and ear cavity are larger at higher fields. Considerable signal is lost there, espectally in EPI. There are also more practical problems. i.e., larger difficulty in making rf volume coils with homogenous sensitivity profile and poorer supports from manufacturers. Nonetheless, Yang et al. (‘96), at 7T, succeeded in imaging of a single barrel field from the rat barrel cortex. Menon et al. (‘97), at 4T, succeeded in observation of the ocular dominance column in the human VI. We have installed a 4T human whole body system (Varian) in RIKEN, and have reproduced the results of previous studies in VI, MT, fusiform face region, and motor cortices, To remove the severe ghosts in multi-shot EPI, we have developed a new pulse sequence (HEADS).