 Our proposed SHNN method has been shown to be more effective than existing methods at identifying optimal cut-off frequencies for filters used with the common spatial pattern, CSP, method and motor imagery, MI, based brain-computer interfaces, BCIs. This method uses SYNC net-based hybrid neural networks to automatically filter raw EEG data, followed by a convolutional neural network to learn deep feature representations from the filtered data. These deep features are then fed into a gated recurrent unit, GRU, module to seek sequential relations, before being passed through a fully connected layer for classification. Our SHNN method achieved higher accuracy rates compared to other state-of-the-art methods on two datasets from the BCI Competition 4. This article was authored by Chang Liu, Jing Jing, Ian Daly, and others.