Low-power DWT-based quasi-averaging algorithm and architecture for epileptic seizure detection

A low-power "near-threshold" epileptic seizure detection processor with multiple algorithm programmability

“…The dynamic power is 360 nW. This is almost 80% lower than the wavelet based algorithm [11]. Although the parameter based algorithm do not show higher efficacy than the wavelet based algorithm, the programmability can be utilized to improve the efficacy with negligible change in power consumption.…”

“…The literature is abundant with numerous algorithms for detection of the epileptic seizure [8][9][10][11]. The characteristic of the seizure is a gradual surge in amplitude of the signal in specific frequency bands.…”

“…The event-based algorithm utilizing LFP provides a low power solution for implantable epilepsy prosthesis [10]. However, its efficacy may be debatable over a range of patients [11]. Utilizing the DWT with a quasi-averaging technique helps in providing the programmability and maintain high efficacy as well [11].…”

“…However, its efficacy may be debatable over a range of patients [11]. Utilizing the DWT with a quasi-averaging technique helps in providing the programmability and maintain high efficacy as well [11]. However, it has a large number of redundant computations and hence is wasteful in terms of energy consumption.…”

Ultra Low-Power Algorithm Design for Implantable Devices: Application to Epilepsy Prostheses

“…On the other hand, implementing a discrete wavelet transform operator (DWT) allows for a much finer time and frequency resolution at a lower hardware cost than traditional FFT or STFT operations for large windows of data. Recently, we reported a DWT-based detection algorithm implemented on silicon with multiplier-less filters, that demonstrates for optimal detection performance at a low hardware cost [45,46]. There have also been numerous reports on the utility of using wavelet transforms (both discrete and continuous) for detection epileptiform activity [47][48][49].…”