Hardware-based simplified discrete wavelet transform for detecting high-voltage spindles in neuron signals

This paper presents a hardware-based simplified discrete wavelet transform (Sim-DWT) that can detect high-voltage spindles (HVSs) in neuron signals. The Sim-DWT can operate at 125 kHz for detection within a spectrum of 5-15 Hz for HVS signals. Only 16 sample points are required as the window length for calculation. No multiplier is required in the Sim-DWT algorithm. The Sim-DWT was implemented in field programmable gate arrays, as well as being tested on real recorded neuron signals. The experimental results of the present study indicate that the hardware-based Sim-DWT algorithm can detect HVSs efficiently, and is suitable for implementation in low-power microcontrollers to serve as the digital core in a deep brain stimulator microsystem.