Open Access
ARTICLE
Brain MRI Patient Identification Based on Capsule Network
Shuqiao Liu, Junliang Li, Xiaojie Li*
Chengdu University of Information Technology, Chengdu, 610225, China
* Corresponding Author: Xiaojie Li. Email:
Journal on Internet of Things 2020, 2(4), 135-144. https://doi.org/10.32604/jiot.2020.09797
Received 20 May 2020; Accepted 15 August 2020; Issue published 22 September 2020
Abstract
In the deep learning field, “Capsule” structure aims to overcome the
shortcomings of traditional Convolutional Neural Networks (CNN) which are
difficult to mine the relationship between sibling features. Capsule Net (CapsNet)
is a new type of classification network structure with “Capsule” as network
elements. It uses the “Squashing” algorithm as an activation function and Dynamic
Routing as a network optimization method to achieve better classification
performance. The main problem of the Brain Magnetic Resonance Imaging
(Brain MRI) recognition algorithm is that the difference between Alzheimer’s
disease (AD) image, the Mild Cognitive Impairment (MCI) image, and the
normal image is not significant. It is difficult to achieve excellent results using a
multi-layer CNN. However, CapsNet can be in the case of a shallower network,
which can accommodate more useful feature information for identifying brain
MRI. In this paper, we designed a shallow CapsNet to identify patients with
brain MRI by binary classification. Compared with VGG16, Resnet34,
DenseNet121 and ResNeXt50. Experimental results illustrate that CapsNet is
superior to CNN network in its accuracy and F1-score. The indicators were
86.67% and 83.33%, respectively. Furthermore, we show that the capsule
network shows excellent performance in brain MRI recognition compared with
those popular networks.
Keywords
Cite This Article
S. Liu, J. Li and X. Li, "Brain mri patient identification based on capsule network,"
Journal on Internet of Things, vol. 2, no.4, pp. 135–144, 2020. https://doi.org/10.32604/jiot.2020.09797