Time-scale analysis of motor unit action potentials

Quantitative analysis in clinical electromyography (EMG) is very desirable because it allows a more standardized, sensitive and specific evaluation of the neurophysiological findings, especially for the assessment of neuromus cular disorders. Following the recent development of computer-aided EMG equ ipment, different methodologies in the time domain and frequency domain hav e been followed for quantitative analysis. In this study, the usefulness of the wavelet transform (WT), that provides a linear time-scale representati on is investigated, for describing motor unit action potential (MUAP) morph ology. The motivation behind the use of the WT is that it provides localize d statistical measures (the scalogram) for nonstationary signal analysis. T he following four WT's were investigated in analyzing a total of 800 MUAP's recorded from 12 normal subjects, 15 subjects suffering with motor neuron disease, and 13 from myopathy: Daubechies nifh four and 20 coefficients, Ch ui (CH), and Battle-Lemarie (BL). The results are summarized as follows: I) most of the energy of the MUAP signal is distributed among a small number of well-localized tin time) WT coefficients in the region of the main spike , 2) for MUAP signals, we look to the low-frequency coefficients far captur ing the average waveshape of the MUAP signal over long durations, and we lo ok to the high-frequency coefficients for locating MUAP spike changes, 3) t he Daubechies 4 wavelet, is effective in tracking the transient components of the MUAP signal, 4) the linear spline CH (semiorthogonal) wavelet provid es the best MUAP signal approximation by capturing most of the energy in th e lowest resolution approximation coefficients, and 5) neural network DY (D Y) of Daubechies 4 and BL WT coefficients was in the region of 66%, whereas DY for the empirically determined time domain feature set was 78%. In conc lusion, wavelet analysis provides a new way in describing MUAP morphology i n the time-frequency plane. This method allows for the fast extraction of l ocalized frequency components, which when combined with time domain analysi s into a modular neural network decision support system enhances further th e DY to 82.5% aiding the neurophysiologist in the early and accurate diagno sis of neuromuscular disorders.