Supplemental Digital Content 1
Manuscript: Do intramuscular temperature and fascicle angle affect ultrasound echo intensity values?
Journal: Medicine and Science in Sports and Exercise
Authors: Matheus D. Pinto, Ronei S. Pinto, Kazunori Nosaka, Anthony J. Blazevich Email: [email protected]
Researchers can capture ultrasound scans of muscles in both the transverse and longitudinal planes. However, the effect of transducer plane acquisition on echo intensity is not well established. The aims of this supplemental digital content were to determine the effects of (1) ultrasound transducer acquisition plane (transverse vs. longitudinal) on vastus lateralis echo intensity, and (2) fascicle angle on echo intensity obtained from longitudinal scans. To test aim 1, linear mixed effect models were fitted by restricted maximum likelihood with ‘knee angle’ and ‘transducer plane’, and their interaction effect, used as fixed effects and ‘subjects’ used as random effects. When a significant fixed effect was detected, pairwise comparisons of least squares means were performed with P values adjusted using Tukey method using “emmeans” R package. To test aim 2, the intra- individual relationships between fascicle angle and echo intensity across knee joint angles using repeated-measures Bland-Altman within-subject correlation using “rmcorr” R package and linear-mixed effect model. As described in the manuscript, muscle thickness decreased with knee flexion, and analyses were performed at a constant depth of the region of interest, i.e., analyses of echo intensity from images obtained when the knee was extended (thicker) were obtained using a region of interest with identical depth to images obtained when the knee was flexed. All procedures are fully described in the manuscript.
Effect of transducer plane on echo intensity:
Statistical analyses revealed no significant effect of transducer plane (F(1,184) = 0.30, P = 0.586), however there was a significant interaction between knee angle and
transducer plane (F(1,184) = 8.53, P = 0.004, Figure 1A). Echo intensity was not significantly different between planes with the knee extended (mean difference: -2.96 a.u.
(95% CI: -6.09 – 0.17), P = 0.071) or flexed (mean difference: 2.03 a.u. (95% CI: -1.1 – 5.16, P = 0.337), however the increases in echo intensity from knee extension to flexion were statistically different (mean difference: -4.99 a.u. (95% CI: -8.36 – -1.62), P = 0.0039) between transverse and longitudinal planes (13.4 ± 5.9% vs. 10.4 ± 7.4%, respectively).
While these small, and possibly practically unimportant, differences were detected, the plane in which the ultrasound images are acquired does not affect the conclusions drawn, that passive flexion of the knee significantly increases echo intensity. These results are presented graphically below (Figure 1).
Figure 1. (A) Interaction plot displaying estimated marginal means (linear prediction of echo intensity) for each level of the factors (Knee Angle; 0° (full extension) and 90° flexion) in each transducer plane. (A) Raw
individual data, indicated by different colours, for echo intensity with the knee in full extension and flexed at 90° in the transverse and longitudinal planes. (B) To facilitate interpretation, echo intensity in the transverse and longitudinal planes are plotted within each knee angle. Dark lines represent the average.
Effect of fascicle angle on vastus lateralis echo intensity
As shown in Figure 2A, a large negative correlation (rrm = -0.87, P < 0.001, 95% CI:
-0.91, -0.81) was found between echo intensity and fascicle angle across joint angles.
Similar results were obtained using analyses of EI using a constant depth of the region of interest (rrm = -0.91, P < 0.001, 95% CI: -0.94, -0.87 – Figure 2B), i.e., constant muscle thickness with the knee extended and flexed.
Figure 2. Intra-individual associations between fascicle angle and echo intensity using within-subject Bland- Altman correlation showing the correlation when echo intensity was calculated (A) using the largest region of interest (ROI) that included as much muscle as possible and (B) at a constant depth of the region of interest in images obtained with the knee extended and flexed. Large negative correlations (rrm = -0.87 and 0.91, for A and B, respectively; P < 0.001) were found between fascicle angle and echo intensity. Three images were obtained for each participant at each joint angle to obtain estimates of echo intensity and fascicle angle, hence each participant has three dots; colors identify participants, and lines show the repeated measures correlation fits for each participant.
There was a significant effect of fascicle angle on echo intensity ((F(1, 15.42) = 55.35, P < 0.001). As shown in Figure 3A, the predicted echo intensity at the mean fascicle angle was 146.27 a.u. (95% CI = 138.8 to 153.8), and the model indicated that echo intensity decreased on average by ~2.0 a.u. (95% CI: -2.5 to -1.5, t = -7.17) for every 1°
increase in fascicle angle.
Further analyses of echo intensity at a constant muscle thickness were performed to control for the significant effect of thickness on echo intensity. There was a significant effect of fascicle angle on echo intensity ((F(1, 16.52) = 101.6, P < 0.001). As shown in Fig 7B, the predicted echo intensity at the mean fascicle angle was 143.55 a.u. (95% CI = 135.4, 151.7), and the model indicated that EI decreased on average by ~2.5 a.u. (95% CI: -2.95 to 2.0, t = -10.1) for every 1° increase in fascicle angle.
Figure 3. Fascicle angle centered around the mean against the echo intensity. Black dots indicate raw data whereas red dots and shading represent marginal effects (predicted echo intensity) and their upper and lower limits of 95% confidence intervals, respectively. Vastus lateralis echo intensity was calculated (A) using the largest region of interest (ROI) that included as much muscle as possible and (B) at a constant depth of the region of interest for both images obtained with the knee extended and flexed (see manuscript for details).