Comparative Study of Deep Learning Models for Segmentation of Corpus Callosum

A vital part of the Brain, which is responsible for the transmission of neural messages between the two hemispheres in the Brain is the Corpus Callosum. Many of the neurodegenerative diseases are related to the morphological properties of the Corpus Callosum. Therefore, its study and analysis become an essential part for the detection of such diseases. Examination of the Magnetic Resonance Images (MRI) through the mid-sagittal plane portrays their structure in the most distinguished manner. This paper carries out a comparative study of three deep learning models such as CE-Net, UNet++ & MultiResUNet for the segmentation of Corpus Callosum in the Brain MRI images using the dataset acquired from open source ABIDE platform. CE-Net gave the best dice similarity coefficient score of 0.9311, among all the three Deep Learning models. Thus, the CE-Net segmentation model can be further used for the classification of neurological disorders.

[1]  Letícia Rittner,et al.  Corpus callosum parcellation methods: a quantitative comparative study , 2018, Medical Imaging.

[2]  Taghi M. Khoshgoftaar,et al.  A survey on Image Data Augmentation for Deep Learning , 2019, Journal of Big Data.

[3]  T. Bourgeron,et al.  Neuroanatomical diversity of corpus callosum and brain volume in the Autism Brain Imaging Data Exchange (Abide) project , 2014, bioRxiv.

[4]  Matthew Lai,et al.  Deep Learning for Medical Image Segmentation , 2015, Deep Learning Applications in Medical Imaging.

[5]  Shenghua Gao,et al.  CE-Net: Context Encoder Network for 2D Medical Image Segmentation , 2019, IEEE Transactions on Medical Imaging.

[6]  Mohammad Sohel Rahman,et al.  MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation , 2019, Neural Networks.

[7]  Rizwan Ahmed Khan,et al.  A novel framework for automatic detection of Autism: A study on Corpus Callosum and Intracranial Brain Volume , 2019, ArXiv.

[8]  Sang Won Seo,et al.  Automatic Segmentation of Corpus Callosum in Midsagittal Based on Bayesian Inference Consisting of Sparse Representation Error and Multi-Atlas Voting , 2018, Front. Neurosci..

[9]  Nima Tajbakhsh,et al.  UNet++: A Nested U-Net Architecture for Medical Image Segmentation , 2018, DLMIA/ML-CDS@MICCAI.

[10]  D. Nguyen,et al.  The corpus callosum: white matter or terra incognita. , 2011, The British journal of radiology.

[11]  Weili Yan,et al.  MRI brain images segmentation , 2000, IEEE APCCAS 2000. 2000 IEEE Asia-Pacific Conference on Circuits and Systems. Electronic Communication Systems. (Cat. No.00EX394).

[12]  Mrugendrasinh Rahevar,et al.  Survey on semantic image segmentation techniques , 2017, 2017 International Conference on Intelligent Sustainable Systems (ICISS).

[13]  Ruchi D. Deshmukh.  Study of Different Brain Tumor MRI Image Segmentation Techniques , 2014 .

[14]  Wojciech Czarnecki,et al.  On Loss Functions for Deep Neural Networks in Classification , 2017, ArXiv.

[15]  Huiquan Wang,et al.  AUTOMATED CORPUS CALLOSUM SEGMENTATION IN MIDSAGITTAL BRAIN MR IMAGES , 2017 .

[16]  Shrish Verma,et al.  Corpus Callosum Segmentation from Brain MRI and its Possible Application in Detection of Diseases , 2019, 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT).

[17]  Srikantan S. Nagarajan,et al.  The Role of Corpus Callosum Development in Functional Connectivity and Cognitive Processing , 2012, PloS one.

[18]  Yili Fu,et al.  Automatic Extraction of the Centerline of Corpus Callosum from Segmented Mid-Sagittal MR Images , 2018, Comput. Math. Methods Medicine.