EEG is an important modality for many medical purposes. However, the low-amplitude of signals (10-to-100μV) and large number of channels (~20) raise numerous challenges, including electrode setup (correct placement, skin preparation, sanitation), patient comfort (number of channels, skin abrasion), and robust acquisition (electrode/wire motion artifacts, wire stray coupling). The recent emergence of low-cost, single-use, flexible, pre-gelled electrode arrays, as in Fig. 16.4.1, delivers significant advantages [1]. Today, these are passive, requiring connection to external readout electronics via a many-channel cable. We present the system in Fig. 16.4.1, having similar flexible form factor, but with the following enhancements: (1) embedded low-noise chopper-stabilized amplifiers using amorphous-silicon (a-Si) thin-film transistors (TFTs) compatible with flexible substrates (i.e. low-temperature-processed, <;180°C); (2) compressive-sensing acquisition and multiplexing of >20 EEG channels onto a single interface using TFT scanning circuits, to substantially ease connection with an embedded IC; and (3) an algorithm whereby spectral-energy features, a generic EEG biomarker, are derived directly from the compressed signals (by a conventional CMOS IC). Seizure detection from the extracted features is demonstrated via analog replay of patient EEG through the system.
[1]
S. Kirolos,et al.
Analog-to-Information Conversion via Random Demodulation
,
2006,
2006 IEEE Dallas/CAS Workshop on Design, Applications, Integration and Software.
[2]
Naveen Verma,et al.
Signal Processing With Direct Computations on Compressively Sensed Data
,
2015,
IEEE Transactions on Very Large Scale Integration (VLSI) Systems.
[3]
Naveen Verma,et al.
Thin-film circuits for scalable interfacing between large-area electronics and CMOS ICs
,
2014,
72nd Device Research Conference.
[4]
Takao Someya,et al.
1 $\mu$m-Thickness Ultra-Flexible and High Electrode-Density Surface Electromyogram Measurement Sheet With 2 V Organic Transistors for Prosthetic Hand Control
,
2014,
IEEE Transactions on Biomedical Circuits and Systems.