Acoustic myography is the recording of sounds produced by contracting muscle. These sounds become louder with increasing force of contraction. We have compared muscle sounds with surface EMG to monitor the dissociation of electrical from mechanical events (presumably, the loss of excitation-contraction coupling) which occur with motor unit fatigue. Acoustic signals were amplified using a standard phonocardiograph, recorded on FM magnetic tape, and digitally analyzed. Muscles were examined at rest, with intermittent contractions, and with sustained contractions. We found that with fatigue, the acoustic amplitude decayed, but the surface EMG amplitude did not. With decreased effort, however, the acoustic and the surface EMG amplitudes declined simultaneously. By simultaneously recording acoustic signals and needle EMG, individual motor units were resolved acoustically in two muscles with decreased numbers of motor units and increased motor unit size. Fasciculations also produced acoustic signals, although no acoustic signal has yet been found that correlates with fibrillations. Analysis of acoustic signals from muscle provides a noninvasive method for monitoring motor unit fatigue in vivo. It may also be useful in distinguishing muscle fatigue from decreased volition.

Acoustic, force, and compound muscle action-potential signals were recorded simultaneously during maximal isometric twitches of frog gastrocnemius muscles. The onset of sound production occurred after the onset of muscle depolarization but before the onset of external force production. Acoustic waveforms consisted of oscillations that initially increased in amplitude, followed by decaying oscillations. The peak-to-peak acoustic amplitude increased with increasing temperature with a Q10 of 2.6 +/- 0.2 over a range of 7.0-25.0 degrees C. The acoustic amplitude increased with increasing muscle length up to approximately 90% of the optimal length for force generation. As length was increased further, the acoustic amplitude decreased. Microphones positioned on opposite sides of the muscle recorded acoustic signals that were 180 degrees out of phase. These results provided evidence that sound production is produced by lateral oscillations of muscle. The oscillation frequency may provide a measure of mechanical properties of muscle.

The changes in mechanical resonant frequency of whole muscles during twitch and tetanic contractions were compared to changes in frequency components of the pressure wave produced by muscles during contraction. Resonant frequencies were determined by imposing sinusoidal length changes on a muscle and observing transverse standing waves when the frequency of length change matched the muscle's resonant frequency or a harmonic of the resonant frequency. Acoustic signal instantaneous frequency spectrums were calculated using time-frequency transformations including the Wigner transform and the exponential distribution. During a tetanic muscle contraction, the peak instantaneous frequency initially increased and then became constant as the force plateau was reached. The resonant frequencies determined during the force plateau and during relaxation spanned the same range as the peak instantaneous frequency of the acoustic signal. These results suggest that the acoustic signal may be useful as a non-invasive monitor of muscle resonant frequency during contraction.

The ratio of acoustic myography (AMG) amplitude to surface electromyography (EMG) amplitude is proposed as a measure of mechanical output compared with electrical activity of the contractile system. AMG to EMG ratios were measured from 16 children with muscle disease diagnosed by clinical criteria, EMG, and/or muscle biopsy. These were compared with the ratios from 11 normal volunteers spanning the same age range (7-16 years). AMG to EMG ratios were significantly (P less than 0.01) different for the two populations. Using a linear discriminant function to define the normal range for AMG to EMG ratios yielded a sensitivity of 82% (13 of 16 abnormals diagnosed) and a specificity of 91% (10 of 11 normals). These findings suggest that surface recordings may provide significant diagnostic information in muscle disease. The accuracy may be improved further by using additional muscles (e.g., paraspinals) and evoked twitches.

No abstract