A Model-Based Approach to the Quantitative Analysis of Eyeblink EMG Response Components

Abstract

Abstract A recently developed method of fragmentary decomposition (FD) of nonstationary physiological signals was extended to eyeblink EMG measurement to quantify all significant stimulus-induced components and identify their parameters. FD provides a representation of single-trial eyeblink EMG as a nonstationary signal with the generic mass potential (GMP) being the universal functional element. The current study uses this model-based signal-processing methodology to identify distinct stimulus-induced components of eyeblink EMG and to extract additional psychophysiological information from the eyeblink signal. Analysis of the single-trial eyeblink EMG records from 10 normal subjects showed that GMP is an adequate functional element of which an eyeblink EMG response is composed. In particular, both spontaneous and stimulus-induced single components of eyeblink EMG are produced by functionally similar mechanisms. However, we found that about 54% of GMPs are combined into complex patterns that respond differently to various experimental conditions. To typify characteristic patterns of eyeblink EMG component composition, we defined two fundamental categories of components: Complex components (CC), comprised of multiple subcomponents (GMPs), versus monolithic components (MC), involving a single GMP. Given the nonstationary character of eyeblink EMG, the stimulus-related appearance of some specific component patterns, such as MCs and CCs, is in essence a probabilistic problem. To characterize the probabilistic structure of eyeblink EMG, we introduce the stimulus dependent probability diagram (SDPD), which shows the probability of appearance of defined component patterns of EMG activity at different times after the stimulus presentation. SDPD analysis shows that the stimulus elicits strong though short-term (phasic) effects on monolithic components and moderate but long-lasting (tonic) effects on complex components.