Power spectrum analysis of compound muscle action potential in carpal tunnel syndrome patients

Journal of Orthopaedic Surgery, Jun 2002 by Ogura, Taku, Kubo, Toshikazu, Okuda, Yoshiki, Lee, Kookho, Et al

ABSTRACT

The objective of using wave-form analysis to assess compound muscle action potential (CMAP) in entrapment neuropathy had not been fully developed. We applied the power spectrum analysis to patients with carpal tunnel syndrome (CTS) for this purpose. 24 patients with CTS were divided into three stages according to Mackinnon's classification, and 50 normal volunteers were examined. CMAP was obtained from the abductor pollicis brevis with supramaximal stimulation to median nerve. Mean and peak frequencies were measured by power spectrum analysis.

The distal latencies of CMAP and the sensory nerve conduction velocities showed some prolongation in CTS patients. Integral values of CMAP were also decreased in CTS patients. Mean and peak frequencies of power spectrum of CMAP in volunteers were 134Hz and 98 Hz, respectively. These values shifted into lower frequencies in CTS patients, namely 102Hz and 61 Hz.

Regardless of clinical stage, distal latency of CTS patients correlated with mean frequency

Key words. power spectrum analysis, compound muscle action potential, carpal tunnel syndrome

INTRODUCTION

When motor nerves are electrically stimulated, the muscles controlled them almost synchronously exhibit compound muscle action potentials (CMAP).1,27 However, the conduction velocity of individual fibers 12, 21 in the nerve trunk differs, and this temporal dispersion comprises CMAP. Generally, applying Fourier transformation in compound waves, such as waves in electromyograms (EMG), any waveform can be decomposed into sine curves with different frequencies,2,13 and waveforms on EMG can almost be regarded as an aggregate of sine curves. Thus, analysis of the waveform frequency is used mainly for surface EMG during voluntary contraction, to quantitatively analyze muscles17, 23, 29 or muscle fatigue.15,24

In a recent study, we assumed CMAP to be an average non-synchronously induced single fiber potential.3, 9, 26 We also analyzed the frequency for changes in the waveform 16, 25 of CMAP during peripheral nerve stimulation in patients with carpal tunnel syndrome (CTS)19,22 which is a typical disease of entrapment neuropathy, and compared them with the waveforms of CMAP in healthy volunteers. And we considered the possibility of clinical diagnosis of power spectrum analysis.

MATERIALS AND METHODS

The subjects were 50 healthy volunteers and 24 patients with CTS. The patients comprised 12 patients with mild symptoms, 5 patients with moderate symptoms and 7 patients with severe symptoms. The severity of disorder was divided broadly into the following three categories according to the Mackinnon's classification14: mild cases were defined as those with intermittent symptoms with a positive Phalen's sign or Tinel's sign; moderate cases as those with decreased muscle strength and a positive Phalen's sign or Tinel's sign without muscular atrophy in the thenar muscles; severe cases as those with muscular atrophy and abnormal findings of sensory symptoms, static 2PD and moving 2PD. We also examined 5 patients who underwent surgical decompression of the carpal tunnel.

The potential was recorded as follows: a surface electrode was placed on the belly of the adductor pollicis brevis at 7 cm proximal to the recording electrode; supramaximum electric stimulation (rectangular waves, duration: 0.2 msec) was applied to the trunk of the median nerve at the wrist, using an amplifier of 50 to 3 kHz to induce CMAP; the obtained potentials were recorded using an electromyogram (model MEM4104, Nihon Kohden Corporation, Tokyo, Japan). We also analyzed the distal latency of CMAP, the integral values, and the frequencies obtained by utilizing the soft ware BIMITUS (Kissei comtec Co., Ltd., Japan), for comparison with those of healthy subjects.

In the analysis of CMAP, the integral value was the sum of the negative and positive portions of a CMAP, with the baseline as the standard. Frequency was analyzed for the whole frequency band up to 2,500 Hz at a sampling frequency of 5 kHz and a frequency resolution of 5 Hz. In order to obtain continuous waveforms, the Harming window function was used. A power spectrum was obtained using fast Fourier transformation. The distal latency, sensory nerve conduction velocities (SCV), integral value, frequency of CMAP were compared, and the relationship between distal latency and frequency were evaluated. Statistical significance was accepted at p

Prolongation of the distal latency, which contributes to diagnosis of CTS, was divided into 3 groups: 5 msec or less, 5 to 6 msec, and 6 msec or more, to prepare a box plot of the SCV and the mean frequency. The chart clearly indicated that both SCV and the mean frequency decreased with prolongation of the distal latency. In particular, the mean frequency significantly differed between CTS patients and healthy subjects, even in patients with a prolongation of 5 msec or less (Fig. 3). Both the mean frequency and the peak frequency improved soon after an operation for CTS, with the mean frequency being 128.8 /- 18.6 Hz, and the peak frequency 95.2 /- 19.6 Hz. The differences, however, were not significant, probably due to the small number of cases.