Chinese character writing and decoding method for invasive brain-computer interface

A method for decoding Chinese character writing for an invasive brain-computer interface, comprising the following steps:

• (1) acquiring an original motor neural signal, filtering the signal to obtain a neural signal with a specified frequency band, and further processing the obtained neural signal to obtain an Entire Spiking Activity (ESA) neural signal;
• (2) standardizing the ESA neural signal, truncating motor neural signals of data segments with a set window length, and constructing a training data set;
• (3) in the training data set, according to Chinese character stroke motion signals and a Chinese character stroke break motion signals of Chinese character writing motion signals, dividing corresponding motor neural signals into writing stroke motor neural signals and the writing stroke break motor neural signals, wherein the motor neural signals correspond to the Chinese character writing motion signals;
• (4) using the motor neural signals and the Chinese character writing motion signals as inputs, and training the following three models:
• 1) using the writing stroke motor neural signals and corresponding character stroke motion signals to train a predictor for decoding Chinese character strokes;
• 2) using the writing stroke break motor neural signals and corresponding character stroke break motion signals to train a predictor for decoding Chinese character stroke breaks;
• and 3) using the writing stroke motor neural signals and the writing stroke break motor neural signals to train a state discriminator for distinguishing a writing stroke and a writing stroke break during the Chinese character writing;
• and
• (5) during application, inputting a motor neural signal to be decoded into the state discriminator, inputting the discriminated writing character stroke motor neural signal into the predictor for decoding Chinese character strokes, and inputting the discriminated writing character stroke break motor neural signal into the predictor for decoding Chinese character stroke breaks;
• and combining the decoded stroke motion signal and the decoded stroke break motion signal to obtain a complete decoded Chinese character motion signal.