18.3 A multi-parameter signal-acquisition SoC for connected personal health applications

Connected personal healthcare, or Telehealth, requires smart, miniature wearable devices that can collect and analyze physiological and environmental parameters during a user's daily routine. To truly support emerging applications (Fig. 18.3.1), a generic platform is needed that can acquire a multitude of sensor modalities and has generic energy-efficient signal processing capabilities. SoC technology gives significant advantages for miniaturization. But meeting low-power, medical grade signal quality, multi-sensor support and generic signal processing is still a challenge. For instance, [1] demonstrated a multi-sensor interface but it lacks support for efficient on-chip signal processing and doesn't have a high performance AFE. [2] showed a very low power signal processor but without support for multi-sensor interfacing. [3] presented a highly integrated SoC but lacking power efficiency. This paper demonstrates a highly integrated low-power SoC with enough flexibility to support many emerging applications. A wide range of sensor modalities are supported including 3-lead ECG and bio-impedance via high-performance and low-power AFE. The ARM Cortex™ M0 processor and matrix-multiply-accumulate accelerator can execute numerous biomedical signal processing algorithms (e.g. Independent Component Analysis (ICA), Principal Component Analysis (PCA,) CWT, feature extraction/classification, etc.) in an energy efficient way without sacrificing flexibility. The diversity in supported modalities and the generic processing capabilities, all provided in a single-chip low-power solution, make the proposed SoC a key enabler for emerging personal health applications (Fig. 18.3.1).