HAL Id: hal-03170996
https://hal.archives-ouvertes.fr/hal-03170996
Submitted on 16 Mar 2021
Wafa Mbarki a,b,⁎, Moez Bouchouich a c, Sebastien Frizzi c,f, Frederick Tshibasu d, Leila Ben Farhat e, Mounir Sayadi a
a. Laboratoire Signal, Image et Maitrise de l’Energie(SIME), ENSIT, Université de Tunis, Tunis, Tunisie
b. Université de Sousse, ENISO, Tunisia
c. Aix Marseille Univ, Université de Toulon, CNRS, LIS, Toulon, France
d. Cliniques Universitaires de Kinshasa, République Démocratique deu Congo
e. Service Imagerie Mdicale, Mongi Slim Hospital, La Marsa, Tunisia
f. Université de Toulon, Département Génie Biologie-IUT, 83957 La Garde, France
ABSTRACT
Axial Lumbar disc herniation recognition is a difficult task to achieve, due to many challenges such as complex background, noise, blurry image. Lumbar discs are small joints that lie between each two vertebrae (L1-L2, L2-L3, L3-L4, L4-L5 and L5-S1). The segmentation and localization of the different discs are the most important tasks in Computer aided diagnosing of herniation. During the last five years, deep learning based methods have set new standards for many computer vision and pattern recognition research. In this work, our objective is to develop an automatic system based on deep convolutional neural network. This Network processes the input MRI (Magnetic Resonance Imaging) in multiple scales of context and then merges the high-level features to enhance the capability of the network to detect discs from lumbar spine. In this study, we are particularly interested in convolutional neural networks (CNN); it was characterized by a topology similar to a visual cortex of mammals. In fact, these kind of techniques has been applied successfully in many classification problems. In order to recognize herniated lumbar disc in Magnetic Resonance Imaging (MRI), we have chosen to use Convolutional neural networks based on VGG16 architecture. Experiments were carried on our own dataset from Sahloul University Hospital of Sousse. The accuracy achieved of the trained model was 94% which represents a high-performance results by providing state of the art. Our system is very efficient and effective for detecting and diagnosing herniated lumbar disc. Therefore, The contribution of this study includes in:
Firstly, the using of the U-net deep neural network architecture to localize and to detail the location of the herniation. Secondly, the using of the axial view MRI in order to locate exactly the pathological and the normal intervertebral discs.The main objective of this paper is to help radiologists in the diagnosing and treating lumbar herniated disc disease.
Contents lists available at ScienceDirect
Interdisciplinary Neurosurgery
journal homepage: www.elsevier.com/locate/inat