Abonnement à la biblothèque: Guest
Portail numérique Bibliothèque numérique eBooks Revues Références et comptes rendus Collections

MACROSCOPIC MUSCULAR MODELING BASED ON IN VIVO 4D RADIOLOGY

DOI: 10.1615/IntJMultCompEng.2011002392
pages 131-142

Cornelia Kober
Hamburg University of Applied Sciences, Lohbruegger Kirchstr. 65, D-21033, Hamburg, Germany

Britt-Isabelle Berg
University Hospital Basel, Basel, Switzerland

Maike Sturmat
TU Braunschweig, Braunschweig, Germany

Joerg Rieger
University of Frankfurt, Frankfurt, Germany

Luigi Gallo
University of Zurich, Zurich, Switzerland

Markus Boel
Department of Mechanical Engineering, TU Braunschweig, Germany

Martin Mack
University of Frankfurt, Frankfurt, Germany

Hans-Florian Zeilhofer
TU Braunschweig, Braunschweig, Germany

Robert Sader
University of Frankfurt, Frankfurt, Germany


MOTS CLÉS: 4D visualization, dynamic MRI, sequential MRI, upper arm musculature, ankle jointflexion, lower leg musculature, biceps brachii muscle, gastrocnemius muscle, direct volume rendering, 3D reconstruction

Résumé

Because of muscular deformation and movement, standard radiology provides only a snapshot of a probably never recurring situation. The scope of this project is dynamic rendering of muscular structures, starting from 4D radiology, namely, 3D plus time, to macroscopic visualization and simulation based thereon. As full realtime 4D MRI is still beyond the technical possibilities for most human muscles, we follow kind of multilevel approach. The first step is the analysis of muscular tissue of cadaveric preparations where validation can be performed by direct comparison. Second, nearly static, but living muscular tissue is studied based on standard 3D MRI. The first step toward time dependency is an ex post composed series of static MRIs where the muscle goes back to relaxed position between the acquisition steps. This is followed by so-called “quasi-continuous” acquisition where, although not in real time, the muscle does not go back to its original state, but however remains in a stretched position during acquisition. The final goal is full real-time data acquisition. The radiological acquisition is followed by highly detailed image processing, segmentation, and visualization where the deforming muscular tissue is subjected to direct volume rendering with special transfer functions. The applied methods are demonstrated for the flexion of a human ankle joint and deforming human upper arm musculature. The visualization techniques proved to be well suited for capturing dynamics, but additional radiological research is strongly needed. The area of application of 4D modeling ranges from biomechanics to medical diagnosis and therapy of muscular disorders.