MORPHOMETRY OF HUMAN THIGH MUSCLES - DETERMINATION OF FASCICLE ARCHITECTURE BY MAGNETIC-RESONANCE-IMAGING

Citation
Sh. Scott et al., MORPHOMETRY OF HUMAN THIGH MUSCLES - DETERMINATION OF FASCICLE ARCHITECTURE BY MAGNETIC-RESONANCE-IMAGING, Journal of Anatomy, 182, 1993, pp. 249-257
Citations number
20
Categorie Soggetti
Anatomy & Morphology
Journal title
ISSN journal
00218782
Volume
182
Year of publication
1993
Part
2
Pages
249 - 257
Database
ISI
SICI code
0021-8782(1993)182:<249:MOHTM->2.0.ZU;2-V
Abstract
A previous investigation suggested that striation patterns spanning in dividual muscles in longitudinally oriented MR images may represent th e orientation of its fascicles. In this study, we confirmed that these striation patterns could be used to infer fascicle orientation and to compute other architectural features of muscles from MR images. The v olumes of 14 muscles within a cadaveric thigh were shown to be estimat ed accurately from cross-sectional MR images by comparison with direct measures from muscle mass. The angles of striations were measured at several positions within vastus medialis and semimembranosus from sagi ttal and frontal-plane MR images. Mathematical techniques were develop ed to infer the 3-dimensional orientation of fascicles based on these striation angles. The angle of striations in a 3rd oblique plane was s hown to agree with mathematical predictions based on these computed or ientations. The pennation angle, defined as the angle between the fasc icles and the line of action of the muscle, predicted from the MR imag es, was similar to directly measured values. Interestingly, the pennat ion angle of these fascicles varied along the length of the muscle; in vastus medialis, pennation angle ranged from 5-degrees to 50-degrees in a proximodistal direction. Procedures were developed and validated to compute fascicle length by projection of fascicle orientation acros s the 3D shape of the muscles. The use of MR images to estimate muscle morphometry could improve greatly the predictive capabilities of musc uloskeletal modelling by reducing the number of unknown model paramete rs.