[ 31 May 2020 | vol. 13 | no. 1 | pp. 17-28 ]

About Authors:

Soraya Nasser1, Moulkheir Naoui2, Ghalem Belalem3 and Said Mahmoudi4
-1Soraya Nasser Department of Computer Science, Université Oran1, Algeria
-2Moulkheir Naoui Department of Computer Science, Université Oran1, Algeria
-3Ghalem Belalem Department of Computer Science, Université Oran1, Algeria
-4Said Mahmoudi Department of Computer Science, Mons University, Belgium


The hippocampus is a complex structure of the bilateral brain located in the temporal lobe below the surface of the cortex, its function is indispensable in cognitive processes, it is one of the first structures affected in Alzheimer's disease, epilepsy and schizophrenia by volumetric atrophy detected in their layers and this can be measured by segmentation from magnetic resonance images (MRI); in this paper we present a method to segment the bilateral hippocampus using a deep-learning model. Deep convolutional neural networks (CNN) have shown great success in recent years, due to their ability to learn meaningful features from a mass of training data. Our method use multiple cohorts (NeuroImaging Tools & Resources Collaboratory (NITRC)) annotated by ITK-SNAP software. CNN architecture proposed in this paper consists of two paths (encoding and decoding ) like U-net without concatenation layer. The results obtained are encouraging.


Hippocampus, Segmentation, Convolutional neural network(CNN), Training, Upconvolution


About this Article: