Advanced muscle imaging in adolescent elite rowers utilizing diffusion tensor imaging: Association of imaging findings with stroke typology

Abstract

Purpose: Muscular overuse injuries are a common health issue in elite athletes. Changes in the muscular microenvironment can be depicted by Diffusion Tensor Imaging (DTI). We hypothesize that the biomechanics of different stroke typologies plays a role in muscle injury and tested our hypothesis by magnetic resonance imaging (MRI) examination of the lumbar spine muscles of adolescent rowers utilizing DTI.

Methods and materials: Twenty-two male elite rowers (12 sweep, 10 scull rowers) with a mean age of 15.8 ± 1.2 years underwent 3-Tesla MRI of the lumbar spine 6 hours after cessation of training. Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) were calculated for the erector spinae and multifidus muscle. Student's t-test was used to test differences of DTI parameters between sweep and scull rowers and a Pearson correlation was utilized to correlate the parameters to training volume.

Results: ADC values in the erector spinae and multifidus muscle were significantly higher (p = 0.039) and FA values significantly lower (p < 0.001) in sweep rowers compared to scull rowers. There was no significant association between DTI parameters and training volume (r ≤ -0.459, p ≥ 0.074).

Conclusions: Our DTI results show that lumbar spine muscle diffusivity is higher in sweep rowers than in scull rowers. Altered muscle diffusivity is suggestive of microscopic tissue disruption and might be attributable to biomechanical differences between stroke typologies.

Fig 1. Magnetic resonance imaging analysis.

Qualitative image analyses were performed on fat-suppressed T2-weighted images (a) and T1 Dixon images (c). Regions of interest (ROIs) were seeded from the upper vertebral endplate level of the L4 vertebra to the lower endplate level of the L5 vertebra (b). Exemplary ROI seeding on fat-only T1 Dixon images in the erector spinae muscle (arrowheads) and multifidus muscle (arrows) (c). ROIs were then transferred to the corresponding tensor maps (d).

Fig 2. ADC and FA values for sweep and scull rowers.

ADC values (a) were significantly higher and FA values (b) significantly lower in sweep rowers. ADC, Apparent Diffusion Coefficient; FA, Fractional Anisotropy.

Fig 3. Color-coded ADC maps with corresponding fat-suppressed T2-weighted images.

Exemplary color-coded ADC maps and T2-weighted inversion-recovery images of the lumbar spine of a 14-year old male scull rower (a,c) and a 15-year old male sweep rower (b,d). ADC values were significantly higher in sweep rowers. No pathological findings were evident on conventional images. ADC Apparent Diffusion Coefficient.

Fig 4. Correlation of DTI parameters with training volume.

Scatter plots showing the correlation results of ADC values (a) and FA values (b) with training volume. There was no significant correlation for either DTI parameter. ADC, apparent diffusion coefficient; DTI, Diffusion Tensor Imaging; FA, fractional anisotropy.