Infant Brain Age Classification: 2D CNN Outperforms 3D CNN in Small Dataset

被引:0
作者
Shabanian, Mahdieh [1 ]
Wenzel, Markus [2 ]
DeVincenzo, John P. [3 ]
机构
[1] Univ Tennessee, Ctr Hlth Sci, Dept Biomed Engn, Memphis, TN 38163 USA
[2] Fraunhofer Inst Digital Med MEVIS, Bremen, Germany
[3] Enanta Pharmaceut, Watertown, MA USA
来源
MEDICAL IMAGING 2022: IMAGE PROCESSING | 2022年 / 12032卷
关键词
Deep learning; CNN; neurodevelopmental; age estimation; infant diseases; MACHINE; MRI;
D O I
10.1117/12.2612887
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Determining if the brain is developing normally is a key component of pediatric neuroradiology and neurology. Brain magnetic resonance imaging (MRI) of infants demonstrates a specific pattern of development beyond simply myelination. While radiologists have used myelination patterns, brain morphology and size characteristics to determine age-adequate brain maturity, this requires years of experience in pediatric neuroradiology. With no standardized criteria, visual estimation of the structural maturity of the brain from MRI before three years of age remains dominated by inter-observer and intra-observer variability. A more objective estimation of brain developmental age could help physicians identify many neurodevelopmental conditions and diseases earlier and more reliably. Such data, however, is naturally hard to obtain, and the observer ground truth not much of a gold standard due to subjectivity of assessment. In this light, we explore the general feasibility to tackle this task, and the utility of different approaches, including two- and three-dimensional convolutional neural networks (CNN) that were trained on a fusion of T1-weighted, T2-weighted, and proton density (PD) weighted sequences from 84 individual subjects divided into four age groups from birth to 3 years of age. In the best performing approach, we achieved an accuracy of 0.90 [95% CI:0.86-0.94] using a 2D CNN on a central axial thick slab. We discuss the comparison to 3D networks and show how the performance compares to the use of only one sequence (T1w). In conclusion, despite the theoretical superiority of 3D CNN approaches, in limited-data situations, such approaches are inferior to simpler architectures. The code can be found in https://github.com/shabanian2018/Age_MRI-Classification
引用
收藏
页数:8
相关论文
共 50 条
  • [11] Multi-Center Brain Imaging Classification Using a Novel 3D CNN Approach
    Yuan, Lin
    Wei, Xue
    Shen, Hui
    Zeng, Ling-Li
    Hu, Dewen
    IEEE ACCESS, 2018, 6 : 49925 - 49934
  • [12] 3D Model Classification and Retrieval Based on CNN and Voting Scheme
    Bai J.
    Si Q.
    Qin F.
    Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics, 2019, 31 (02): : 303 - 314
  • [13] An efficient implementation of 2D convolution in CNN
    Chang, Jing
    Sha, Jin
    IEICE ELECTRONICS EXPRESS, 2017, 14 (01): : 1 - 8
  • [14] Autofocus Net: Auto-focused 3D CNN for Brain Tumour Segmentation
    Stefani, Andreas
    Rahmat, Roushanak
    Harris-Birtill, David
    MEDICAL IMAGE UNDERSTANDING AND ANALYSIS, 2020, 1248 : 43 - 55
  • [15] Face recognition in unconstrained environments with multimodal 2D/3D BiLSTM-CNN parallel model
    Bahroun, Sahbi
    SIGNAL IMAGE AND VIDEO PROCESSING, 2025, 19 (07)
  • [16] A Simple Methodology for 2D Reconstruction Using a CNN Model
    Levid Rodriguez-Santiago, Armando
    Anibal Arias-Aguilar, Jose
    Elias Petrilli-Barcelo, Alberto
    Miranda-Luna, Rosebet
    PATTERN RECOGNITION (MCPR 2020), 2020, 12088 : 98 - 107
  • [17] 3D CNN for Human Action Recognition
    Boualia, Sameh Neili
    Ben Amara, Najoua Essoukri
    2021 18TH INTERNATIONAL MULTI-CONFERENCE ON SYSTEMS, SIGNALS & DEVICES (SSD), 2021, : 276 - 282
  • [18] A Multichannel Hybrid 2D-3D-CNN for Hyperspectral Image Classification With Small Training Sample Sizes
    Chen, Shih-Yu
    Chu, Po-Yu
    Liu, Kuan-Liang
    Wu, Yu-Cheng
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62
  • [19] An Enhanced Spectral Fusion 3D CNN Model for Hyperspectral Image Classification
    Zhou, Junbo
    Zeng, Shan
    Xiao, Zuyin
    Zhou, Jinbo
    Li, Hao
    Kang, Zhen
    REMOTE SENSING, 2022, 14 (21)
  • [20] Estimating Biological Age from Physical Activity using Deep Learning with 3D CNN
    Rahman, Syed Ashiqur
    Adjeroh, Donald A.
    2019 IEEE INTERNATIONAL CONFERENCE ON BIOINFORMATICS AND BIOMEDICINE (BIBM), 2019, : 1100 - 1103