Purpose To evaluate the reliability and validity of objective fat quantification using the modified Dixon magnetic resonance imaging sequence for assessing fatty degeneration in conservatively treated patients with proximal hamstring avulsion injuries. Methods Magnetic resonance imaging with modified Dixon sequence was used. The fat percentage was measured in the modified Dixon FatFraction Map by setting a region of interest, whereby 3 methods were used with decreasing accuracy: anatomical outline of the entire bundle region of interest, respective elliptical region of interest of all 3 muscle bellies, and individual elliptical regions of interest of the bundle. In addition, the ischiocrural muscle cross‐sectional area and the Goutallier‐like classification were determined as a measure of atrophy. Results Magnetic resonance imaging was performed 35.8 ± 22.43 months after injury. Data from 16 patients (11 women, 5 men) with conservatively treated, unilateral proximal tendon rupture of the ischiocrural muscle (Wood type 3‐5) were included. When determining the cross‐sectional area and the modified Dixon fat content, the inter‐ and intraobserver reliability was good (all ≥ 0.933). In comparison, the reliability values of the Goutallier‐like classification were worse (0.897 and 0.733). The entire bundle and muscle belly region of interest methods for determining the modified Dixon fat content showed comparable results and are therefore suitable for fat quantification. The fat content determined in the anatomical region of interest correlated significantly with the stages of the Goutallier‐like classification (0.823, 95% confidence interval: 0.659, 0.912, P < .001) and the muscle cross‐sectional area (0.363, 95% confidence interval: −0.005, 0.638, P = .041). Conclusions The modified Dixon sequence represents a quantitative‐objective measure and is suitable as a measurement tool for fatty atrophy after proximal tendon injuries of the ischiocrural muscles. A simplified region of interest placement approach showed good validity and reproducibility. Level of Evidence Level III, retrospective diagnostic comparative study.
Hanke et al. (Mon,) studied this question.