Contents I. Introduction
Nowadays, bidirectional neuromodulation technologies with simultaneous recording and stimulation are urgently demanded in both neuroscience research and neurobiological disease treatment, such as brain-machine-interface (BMI) experiments, stroke rehabilitation, Parkinson’s disease, epilepsy, and severe depression [1]–[11]. Some recent neural recording experiments show that extracellular spikes are from several V to several hundreds of V, local field potentials (LFPs) or power line interferences are at a range of mV, and motion artifacts reach tens of mV [12]–[14]. Thus, a wide system dynamic range is required to simultaneously record LFPs, extracellular spikes, motion artifacts, and power line interferences without saturating circuits. The requirement on system dynamic range is further pushed when supporting more sophisticated neuroscience experiments and clinical applications, where electrical microstimulation is used for probing neural circuitry and identifying networks of neurons. However, large amplitude stimulation artifacts (>100 mV) that frequently appear in the experimentsapplications [15], [16] will saturate the recording systems. Thus, one key challenge is how to reduce recovery time from system saturation due to electrical microstimulation.
