Project-Computational-Neuroscience

This repository is about project of computational neuroscience

1.Problem statement

Decoding arm motion using electrocorticographic (ECoG) signals in monkeys using CNN, LSTM, and PLS.

2.Related works

explores existing research and solutions related to your project

3.Proposed Method

Feature extraction using Morlet wavelet transform and decoding with neural networks and PLS.

4.Implementation

4.1. Dataset

Under this subsection, you'll find information about the dataset used for your project. It includes details about the dataset source,task, channels, and number of trials. http://neurotycho.org/expdatalist/listview?task=36 ECoG signals and arm motion were recorded during a food tracking task in two Japanese macaques. Signals from Monkey A were recorded using 32 electrodes, and signals from Monkey K were recorded using 64 electrodes. The two monkeys together performed 35 trials.

4.2. Model

We used LSTM and CNN as deep learning models, and PLS as a classical regression model for decoding. Morlet wavelet transform was used for feature extraction.

4.3. Evaluate

To evaluate the efficiency of the decoding model, we calculated the correlation coefficient between the predicted and observed trajectories.