Fully automated segmentation of the left ventricle in cine cardiac MRI using neural network regression

被引：59
|
作者：
Tan, Li Kuo [1 ,2 ]
McLaughlin, Robert A. [3 ,4 ]
Lim, Einly [5 ]
Aziz, Yang Faridah Abdul [1 ,2 ]
Liew, Yih Miin [5 ]
机构：
[1] Univ Malaya, Dept Biomed Imaging, Fac Med, Kuala Lumpur, Malaysia
[2] Univ Malaya, Res Imaging Ctr, Kuala Lumpur, Malaysia
[3] Univ Adelaide, Adelaide Med Sch, Fac Hlth & Med Sci, Australian Res Council Ctr Excellence Nanoscale B, Adelaide, SA, Australia
[4] Univ Adelaide, Inst Photon & Adv Sensing IPAS, Adelaide, SA, Australia
[5] Univ Malaya, Dept Biomed Engn, Fac Engn, Kuala Lumpur, Malaysia
来源：
JOURNAL OF MAGNETIC RESONANCE IMAGING | 2018年 / 48卷 / 01期
关键词：
cardiac MRI; LV segmentation; deep learning; cine MRI; automated segmentation; QUANTIFICATION; CONSENSUS; MASS;
D O I：
10.1002/jmri.25932
中图分类号：
R8 [特种医学]; R445 [影像诊断学];
学科分类号：
1002 ; 100207 ; 1009 ;
摘要：
BackgroundLeft ventricle (LV) structure and functions are the primary assessment performed in most clinical cardiac MRI protocols. Fully automated LV segmentation might improve the efficiency and reproducibility of clinical assessment. PurposeTo develop and validate a fully automated neural network regression-based algorithm for segmentation of the LV in cardiac MRI, with full coverage from apex to base across all cardiac phases, utilizing both short axis (SA) and long axis (LA) scans. Study TypeCross-sectional survey; diagnostic accuracy. SubjectsIn all, 200 subjects with coronary artery diseases and regional wall motion abnormalities from the public 2011 Left Ventricle Segmentation Challenge (LVSC) database; 1140 subjects with a mix of normal and abnormal cardiac functions from the public Kaggle Second Annual Data Science Bowl database. Field Strength/Sequence1.5T, steady-state free precession. AssessmentReference standard data generated by experienced cardiac radiologists. Quantitative measurement and comparison via Jaccard and Dice index, modified Hausdorff distance (MHD), and blood volume. Statistical TestsPaired t-tests compared to previous work. ResultsTested against the LVSC database, we obtained 0.770.11 (Jaccard index) and 1.330.71mm (MHD), both metrics demonstrating statistically significant improvement (P < 0.001) compared to previous work. Tested against the Kaggle database, the signed difference in evaluated blood volume was +7.213.0mL and -19.818.8mL for the end-systolic (ES) and end-diastolic (ED) phases, respectively, with a statistically significant improvement (P < 0.001) for the ED phase. Data ConclusionA fully automated LV segmentation algorithm was developed and validated against a diverse set of cardiac cine MRI data sourced from multiple imaging centers and scanner types. The strong performance overall is suggestive of practical clinical utility. Level of Evidence: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018.
引用
下载
收藏
页码：140 / 152
页数：13
相关论文
共 50 条
  • [11] Segmentation of the left ventricle in cardiac cine MRI using a shape-constrained snake model
    Wu, Yuwei
    Wang, Yuanquan
    Jia, Yunde
    COMPUTER VISION AND IMAGE UNDERSTANDING, 2013, 117 (09) : 990 - 1003
  • [12] Fully automated cardiac MRI segmentation using dilated residual network
    Ahmad, Faizan
    Hou, Wenguo
    Xiong, Jing
    Xia, Zeyang
    MEDICAL PHYSICS, 2023, 50 (04) : 2162 - 2175
  • [13] Watershed segmentation of basal left ventricle for quantitation of cine cardiac MRI function
    Yingli Lu
    Kim A Connelly
    Alexander J Dick
    Graham A Wright
    Perry E Radau
    Journal of Cardiovascular Magnetic Resonance, 13 (Suppl 1)
  • [14] Accurate Segmentation Framework for the Left Ventricle Wall from Cardiac Cine MRI
    Sliman, H.
    Khalifa, F.
    Elnakib, A.
    Soliman, A.
    Beache, G. M.
    Gimel'farb, G.
    Emam, A.
    Elmaghraby, A.
    El-Baz, A.
    2013 INTERNATIONAL SYMPOSIUM ON COMPUTATIONAL MODELS FOR LIFE SCIENCES, 2013, 1559 : 287 - 296
  • [15] Left Ventricle Segmentation in Cardiac MR Images Using Fully Convolutional Network
    Nasr-Esfahani, Mina
    Mohrekesh, Majid
    Akbari, Mojtaba
    Soroushmehr, S. M. Reza
    Nasr-Esfahani, Ebrahim
    Karimi, Nader
    Samavi, Shadrokh
    Najarian, Kayvan
    2018 40TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY (EMBC), 2018, : 1275 - 1278
  • [16] Automated Segmentation of Cardiac Chambers from Cine Cardiac MRI Using an Adversarial Network Architecture
    Upendra, Roshan Reddy
    Dangi, Shusil
    Linte, Cristian A.
    MEDICAL IMAGING 2020: IMAGE-GUIDED PROCEDURES, ROBOTIC INTERVENTIONS, AND MODELING, 2021, 11315
  • [17] Robust Segmentation of the Left Ventricle from Cardiac MRI via Capsule Neural Network
    Dong, Jiaxiu
    Liu, Chang
    Yang, Cong
    Lin, Nan
    Cao, Yangjie
    ISICDM 2018: PROCEEDINGS OF THE 2ND INTERNATIONAL SYMPOSIUM ON IMAGE COMPUTING AND DIGITAL MEDICINE, 2018, : 88 - 91
  • [18] Segmentation of Left Ventricle Myocardium in Porcine Cardiac Cine MR Images Using a Hybrid of Fully Convolutional Neural Networks and Convolutional LSTM
    Zhang, Dongqing
    Icke, Ilknur
    Dogdas, Belma
    Parimal, Sarayu
    Sampath, Smita
    Forbes, Joseph
    Bagchi, Ansuman
    Chin, Chih-Liang
    Chen, Antong
    MEDICAL IMAGING 2018: IMAGE PROCESSING, 2018, 10574
  • [19] A Robust, Automated Left Ventricle Region of Interest Localization Technique using a Cardiac Cine MRI Atlas
    Ben-Zikri, Yehuda Kfir
    Linte, Cristian A.
    MEDICAL IMAGING 2016: IMAGE-GUIDED PROCEDURES, ROBOTIC INTERVENTIONS, AND MODELING, 2016, 9786
  • [20] Automatic Left Ventricle Segmentation from Short-Axis Cardiac MRI Images Based on Fully Convolutional Neural Network
    Shaaf, Zakarya Farea
    Jamil, Muhammad Mahadi Abdul
    Ambar, Radzi
    Alattab, Ahmed Abdu
    Yahya, Anwar Ali
    Asiri, Yousef
    DIAGNOSTICS, 2022, 12 (02)
12345