Semi-automatic segmentation of intervertebral disc for diagnosing herniation using axial view MRI

Wafa Mbarki , Moez Bouchouicha (1) , Sebastien Frizzi , Frederick Tshibasu , Leila Ben Farhat , Mounir Sayadi

1 LIS - Laboratoire d'Informatique et des Systèmes (LIS) (Marseille, Toulon)

We consider the problem of lower back pain and sciatica due to the loss of the disc's height and the displacement of vertebrae. Our spine represents a combination of discs and vertebrae; between each two vertebrae, we can find an intervertebral disc. We will be interested in this paper to the lumbar discs, which are the most responsible for the lumbar herniation. Computer Aided Diagnosing (CAD) system for localizing herniated and normal intervertebral discs is a difficult task due to the method for treatment. Magnetic Resonance Imaging (MRI) are widely used to diagnose lower back pain and sciatica. We will be concentrated in this work on the T2-axial view MRI to successfully detect and classify the intervertebral discs which are the most important tasks to discuss in a system CAD. The originality of this paper consists in the development of a new method based on active contour and intuitionistic fuzzy C means (IFS) techniques to localize and extract disc from axial view MRI in order to find the type of herniated lumbar disc as foraminal, median or post lateral, we achieved 0.86 dice similarity index on 185 T2 axial MRI.

Mots clés: Axial view MRI; detection; herniated lumbar disc; Acitve Contour; IFS; Axial view MRI

Domaines: Sciences de l'ingénieur [physics]

Wafa Mbarki, Moez Bouchouicha, Sebastien Frizzi, Frederick Tshibasu, Leila Ben Farhat, et al.. Semi-automatic segmentation of intervertebral disc for diagnosing herniation using axial view MRI. International Conference on Advanced Technologies for Signal and Image Processing, Sep 2020, Sfax, Tunisia. ⟨10.1109/ATSIP49331.2020.9231737⟩. ⟨hal-03171620⟩