Convolutional neural network-based regression for quantification of brain characteristics using MRI
Fernandes, João; Alves, Victor; Khalili, Nadieh; Benders, Manon J.N.L.; Išgum, Ivana; Pluim, Josien; Moeskops, Pim
(2019)
New Knowledge in Information Systems and Technologies - Volume 2, volume 931, pp. 577 - 586
Advances in Intelligent Systems and Computing, volume 931, pp. 577 - 586
World Conference on Information Systems and Technologies, WorldCIST 2019, volume 931, pp. 577 - 586
(Part of book)
Abstract
Preterm birth is connected to impairments and altered brain growth. Compared to their term born peers, preterm infants have a higher risk of behavioral and cognitive problems since most part of their brain development is in extra-uterine conditions. This paper presents different deep learning approaches with the objective of quantifying
... read more
the volumes of 8 brain tissues and 5 other image-based descriptors that quantify the state of brain development. Two datasets were used: one with 86 MR brain images of patients around 30 weeks PMA and the other with 153 patients around 40 weeks PMA. Two approaches were evaluated: (1) using the full image as 3D input and (2) using multiple image slices as 3D input, both achieving promising results. A second study, using a dataset of MR brain images of rats, was also performed to assess the performance of this method with other brains. A 2D approach was used to estimate the volumes of 3 rat brain tissues.
show less
Download/Full Text
The full text of this publication is not available.
Keywords: Brain quantification, Convolutional neural networks, Deep learning, Magnetic resonance imaging, Preterm infants, Rat brain, Regression, Control and Systems Engineering, General Computer Science
ISSN: 2194-5357
ISBN: 9783030161835
Publisher: Springer Verlag
Note: Funding Information: This work was supported by COMPETE: POCI-01-0145-FEDER-007043 and FCT – Fundação para a Ciência e Tecnologia within the Project Scope: UID/CEC/00319/2013. We gratefully acknowledge the support of the NVIDIA Corporation with their donation of a Quadro P6000 board used in this research. Funding Information: Acknowledgements. This work was supported by COMPETE: POCI-01-0145-FEDER-007043 and FCT – Fundação para a Ciência e Tecnologia within the Project Scope: UID/CEC/00319/2013. We gratefully acknowledge the support of the NVIDIA Corporation with their donation of a Quadro P6000 board used in this research. Funding Information: The dataset with rats MRI scans was acquired from a part of a project called SIGMA, partially financed by Fundação para a Ciência e Tecnologia and Agence National de Recherche (ref: FCT-ANR/NEU-OSD/0258/2012). Wistar rats were the species of rodents used in this study. Using a 2 ⨯ 2 surface coil associated with software Par-avision 6, the 139 scanning sessions were carried out on a Bruker Biospec 11.7T preclinical scanner. SE-EPI diffusion sensitive acquisitions were used, with Time to Repetition = 5 s, Time to Echo = 20 ms, in-plane resolution of 0.375 ⨯ 0.375 mm, slice thickness of 0.5 mm over 40 slices and a Field-of-View of 24 mm. Each acquisition acquired and averaged 10 volumes. Using WM, GM and CSF priors, SPM Segment was used with the objective of creating the ground truth for segmentation [16]. Publisher Copyright: © Springer Nature Switzerland AG 2019.
(Peer reviewed)