Real-time estimation of phase and amplitude with application to neural data

Abstract Computation of the instantaneous phase and amplitude via Hilbert Transform is a powerful tool data analysis. This approach finds many applications in various science engineering branches but not proper for causal estimation because it requires knowledge signal’s past future. However, several problems require real-time amplitude; an illustrative example phase-locked or amplitude-dependent stimulation neuroscience. In this paper, we discuss compare three algorithms that do rely on exploit well-known physical phenomena, synchronization resonance. After testing synthetic set, illustrate their performance computing accelerometer tremor measurements Parkinsonian patient’s beta-band brain activity.