Location and sequence of muscle onset in deep abdominal muscles measured by different modes of ultrasound imaging

Abstract

Various modes of ultrasound (US) imaging have been introduced as an alternative to electromyography for determining muscle onset. The purpose of this study was to compare the agreement between US motion-mode (US_m-mode) and US strain rate (US_SR) derived from tissue velocity imaging in determining latency time, location and sequence of muscle onset in abdominal muscles using the same data set (contractions). Twenty-four subjects performed four rapid arm flexions in response to a light signal while US recordings were made from the abdominal muscles on the contralateral side. The examined muscles were transversus abdominis (TrA), superficial and deep obliquus internus abdominis (OI_deep and OI_sup), and obliquus externus abdominis (OE). The results showed that the two methods detected the first muscle onset on average within 0.1 ms (95% CI; ±1.4 ms) of each other. US_SR detected the second muscle onset on average 27 ms after US_m-mode. While US_SR and US_m-mode can be used interchangeably to detect the first muscle onset, the location of both first onset and subsequent muscle onsets can be reliably detected by US_SR only. Furthermore, this study indicates that OI may be functionally subdivided into a superficial and deep region, with onset in OI_deep occurring on average 53 ms before OI_sup. First onset was detected more frequently in OI than in TrA (65% versus 25% of detected onsets, 10% were equal).

Introduction

In response to anticipated postural perturbations, several studies have shown that muscle onset in deep abdominal muscles is delayed in low back pain patients compared to healthy controls (Hodges and Richardson, 1996, Hodges and Richardson, 1998). Muscle onset in deep trunk muscles has traditionally been determined by use of intramuscular electromyographic (EMG) recordings. Intramuscular EMG is however less than optimal, particularly in clinical studies with repeated measures, as it is invasive, often painful, the electrodes have a limited pick-up area, and the insertion location is difficult to reproduce.

Ultrasound (US) imaging has been introduced as an alternative to EMG for measuring muscle onset (Grubb et al., 1995, Hodges et al., 2003b, Mannion et al., 2008, Pulkovski et al., 2008). The basis for this approach is that muscle activity creates instantaneous structural deformation which can be detected by US imaging. Motion-mode (US_m-mode) imaging is based on recordings of directional tissue displacements towards and away from the transducer over time, and can thereby detect the temporal starting point of muscle activity (onset). However, US_m-mode cannot accurately determine the site of onset nor subsequent onsets in other muscles along the scan line once the initial onset is triggered (Mannion et al., 2008, Vasseljen et al., 2009). To solve this, tissue velocity imaging (TVI), which quantifies both the direction and the velocity of tissue displacement has been introduced (Grubb et al., 1995, Heimdal et al., 1998, Peolsson et al., 2008). TVI serves as basis for the processing of tissue strain rate (US_SR), i.e., “rate of stretch” (Stoylen et al., 2000), which can detect velocity gradients over small regional depth ranges, giving information on compression and thickening of muscle layers or depth ranges along the US beam axis. US_SR can thus provide information on the location of both the initial and subsequent onsets by studying the start of muscle thickening in the various muscle layers underneath the transducer.

Muscle deformation onset determined by US imaging has shown results in agreement with muscle activity onset determined by intramuscular EMG (McMeeken et al., 2004, Vasseljen et al., 2009). Both US_m-mode (Mannion et al., 2008, Vasseljen et al., 2006) and US_SR (Vasseljen et al., 2009) are shown to be valid and reliable for detecting the first muscle onset in deep abdominal muscles, while subsequent onsets from different muscles or depth layers can reliably be detected by US_SR only. However, the US_SR method is technically challenging causing considerable problems with missing data and interpretation difficulties due to muscle movement occurring transversally to the US beam axis (Marwick, 2006, Vasseljen et al., 2009). These are minor problems with US_m-mode. Missing data is a problem particularly in clinical trials since it threatens a priori sample size estimates and generalizability of the results. If latency time, irrespective of location of onset proves to be the critical parameter in low back pain, then US_m-mode and US_SR may be used interchangeably providing they give similar results. These two US based methods have not been directly compared for onset estimates.

The main purpose of this study was to investigate the agreement between US_m-mode and US_SR in determining first muscle onset in deep abdominal muscles based on the same real-time US tissue velocity recording. Secondary aims were to investigate the location for the initial onset and whether the two methods can be used to differentiate regional differences in subsequent muscle onset.