Applied Physiology

Eur J Appl Physiol (1991) 63:424-429

European Journal of

and Occupational Physiology © Springer-Verlag 1991

Muscle cross-section measurement by magnetic resonance imaging Ralph Beneke ~, JiJrg Neuerburg 2, and Klaus Bohndorf 2 i Institut fiir Sportmedizin, Freie Universit~it Berlin, Oskar-Helene-Heim, Clayallee 229, W-1000 Berlin 33, Federal Republic of Germany 2 Klinik far Radiologische Diagnostik des Klinikums der RWTH Aachen, Pauwelsstral3e 1, W-5100 Aachen, Federal Republic of Germany Accepted August 8, 1991

Summary. Muscle cross-section areas were measured by magnetic resonance imaging (MRI) in the thigh of a human cadaver, the results being compared with those obtained by photography of corresponding anatomic macroslices. A close correlation was found between MRI and photographic evaluation, differences between the methods ranging from nil to 9.5%, depending on the scan position and the muscle groups. In vivo MRI measurements were performed on 12 female and 16 male students, the objectivity, the test-retest reliability and the variability of the MRI measurements being studied by fixing the scan position either manually or by coronary scan. The latter method appeared to be more objective and reliable. The coefficients of variation for muscle cross-section areas measured by MRI were in the range of those for the planimetry of given cross-section areas. Allowing for differentiation between several small muscle bundles in a given area, MRI proved to be a suitable method to quantify muscle cross-sections for intra- and interindividual analysis of muscle size. Key words: Magnetic resonance imaging - Muscle size measurement

3. Isotope methods (Miller and Nelson 1976; Saziorski et al. 1984) 4. Ultrasonography, used as A- or B- or immersion scan (Ikai and Fukunaga 1968, 1970; Fukunaga 1976; Young et al. 1981, 1983, 1984; Young and Hughes 1982; Hicks et al. 1984; Woltering et al. 1987; H~ikkinen and Keskinen 1989) and 5. X-ray computed tomography (Haeggmark et al. 1978; Maughan et al. 1983; Hudasch et al. 1985; de Koenig et al. 1986; Beneke 1988; Beneke et al. 1990). With the exception of ultrasound, none of the above methods can actually be considered as being simultaneously safe, objective, valid and reliable. Single muscles were first identified in vivo by magnetic resonance imaging (MRI) in 1979 (Hinshaw et al. 1979). Since 1988 MRI has been used to quantify crosssection areas of muscle groups for comparing absolute forces (Narici et al. 1988) and revealing adaptations brought about by training (Narici et al. 1989). The present work adapts a standardized protocol, originally developed for x-ray computed tomography (CT), that is suitable for inter- and intra-individual comparison of muscle cross-sections, independent of subcutaneous fat (Beneke 1988). Validity, objectivity and reliability of the MRI method for muscle cross-section area measurement have been analysed.

Introduction Exact determinations of muscle size are required for the analysis of muscle power (Young et al. 1983; Beneke 1988). Methods for muscle size measurement comprise 1. Tests with cadavers and the transfer of results to in vivo conditions (Henke 1865; Koster 1867; Hermann 1898; Franke 1920) 2. Circumference measurements that are or are not combined with standard x-ray techniques (Morris 1948; Stoboy 1984)

Offprint requests to: R. Beneke

Methods A whole-body-coil and a super-conducting magnet (Magnetom, Siemens AG, Erlangen, Germany) with a fieldstrength of 1.5 T were used. Tl-weighted spin-echo sequences were taken by a multislice-procedure (slice thickness 8 ram) with a repetition time of 0.4 s and an echo time of 15 ms. An enlarged picture matrix of 256 x 256 pixel was applied, the time per measurement being 3.4 min. Cross-section measurements in the cadaveric thigh. The validity of MRI cross-section measurements was assessed on a cadaveric thigh, exarticulated at the hip. An axial picture of the thigh was taken proximal to the knee-joint and in the middle of the upper thigh, the scan positions were marked with a pencil on the skin.

425 The following cross-section areas were measured: (a) upper thigh including subcutaneous fat and femur bone, (b) quadriceps femoris muscle, (c) vastus medialis, vastus intermedius quadriceps femoris muscle, (d) vastus lateralis quadriceps femoris muscle, (e) rectus femoris muscle, (f) sartorius muscle and (g) femur bone. Cross-section areas were calculated by software installed in the MRI hardware in the following way. After choosing a region of interest in the axial picture, this area was enlarged on a monitor screen as much as was necessary and subjected to planimetry. Frozen anatomical sections were sliced according to the MRIscan positions and photographed. Cross-section areas were calculated with the MOP analysing system (Fa. Kontron, Eching, FRG). An average of five measurements was taken for each value. The results obtained by the photographing of anatomic macroslices served as a reference for assessment of the relative difference as well as the mean relative error of the prediction by linear regression (Iresiduall × 100 x measured area-l).

Cross-section measurements in the thigh of volunteers. A group of 12 female [25.6 (SD 1.5) years; height 166.7 (SD 2.1) cm; body mass 59.2 (SD 5.9) kg] and 16 male [27.7 (SD 2.6) years; height 182.4 (SD 2.9) cm; body mass 76.0 (SD 5.5) kg] healthy students volunteered for muscle size measurements. First the distance between knee-joint-cleft and spina iliaca anterior superior was measured with a tape measure. The scan position was fixed at onethird of the distance between these two points cranial of the kneejoint-cleft. In 14 volunteers (group 1) the scan position was marked on the skin with a pencil. Then the position for an axial scan was fixed by adjusting the examination table. In 14 volunteers (group 2) a multiplanar coronary scan of the femur bone was conducted first. After marking the knee-joint cleft on this picture the definitive axial scan position was fixed by MRI control as described above. Performing the two methods of positioning on the same person would have been ideal, but this was not possible because of limited availability of volunteers. During the examination, the volunteers' legs were fastened to the table beneath them with belts proximal to the ankle and the knee-joints. For each measurement the tasks necessary to obtain a cross-section area were distributed randomly among three researchers. The complete examination including positioning and measuring took about 20 rain. Each volunteer was examined twice on different occasions. A third measurement was performed on 3 subjects from group 1 and on 4 from group 2. Stat&tical analysis. Intra-individual mean area differences between the first and the second measurement were analysed by paired t-test and Student's t-test. The mean relative error for prediction by linear regression (Iresidualsl x 100 x measured area -1) was computed. Coefficients of variation for planimetry and crosssection area measurement were calculated. Statistical significance was considered to be present at P

Muscle cross-section measurement by magnetic resonance imaging.

Muscle cross-section areas were measured by magnetic resonance imaging (MRI) in the thigh of a human cadaver, the results being compared with those ob...
723KB Sizes 0 Downloads 0 Views