Multiorgan Motion Tracking in Dynamic Magnetic Resonance Imaging for Evaluation of Pelvic System Mobility and Shear Strain

Zhifan Jiang (1) , Jean-François Witz (1, 2) , Pauline Lecomte-Grosbras (1) , Jeremie Dequidt (3, 4) , Stéphane Cotin (5) , Christelle Rubod (6) , Christian Duriez (3) , Mathias Brieu (1, 7)

1 LML - Laboratoire de Mécanique de Lille - FRE 3723
2 CNRS - Centre National de la Recherche Scientifique
3 DEFROST - Deformable Robots Simulation Team
4 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
5 MIMESIS - Computational Anatomy and Simulation for Medicine
6 Service de gynécologie obstétrique
7 Centrale Lille

Female pelvic disorders have a large social impact, the diagnosis of which relies on a key indication: pelvic mobility. The normal mobility is present in a healthy patient, meanwhile the hypermobility can be a sign of female pelvic prolapse and the hypomobility for endometriosis. The evaluation of pelvic mobility is based on medical image analysis. However, the latter does not provide precise values of these indicators directly. Moreover, suspension devices play an important role in pelvic organ function, but can hardly be observed on medical images. Our objective is to propose an image-based analysis tool for the quantitative evaluation of pelvic mobility and the shear strain which has an impact on suspension devices. Hence, this paper introduces a such tool based on an e cient and semi-automatic motion tracking of multiple pelvic organs: the bladder, vagina and rectum presented in dynamic magnetic resonance imaging sequences. The method was validated on prototypical images and applied to di↵erent mobility cases. The computed displacement and shear strain fields provide important information on the quality of suspension devices between organs for a fine diagnosis in the clinical context, for example the early diagnosis of female pelvic prolapse and the localization of possible lesion areas before surgery. Meanwhile, the predicted mobility can be used to compare with the finite element model for numerical simulation.

Mots clés: image correlation; pelvic system mobility; simulation 3D; motion tracking; shear strain

Domaines: Informatique [cs]; Imagerie médicale; Biomécanique [physics.med-ph]

Zhifan Jiang, Jean-François Witz, Pauline Lecomte-Grosbras, Jeremie Dequidt, Stéphane Cotin, et al.. Multiorgan Motion Tracking in Dynamic Magnetic Resonance Imaging for Evaluation of Pelvic System Mobility and Shear Strain. Strain, 2017, 53 (2), ⟨10.1111/str.12224⟩. ⟨hal-01496177v2⟩