Literature DB >> 31729078

Accuracy, uncertainty, and adaptability of automatic myocardial ASL segmentation using deep CNN.

Hung P Do1, Yi Guo1, Andrew J Yoon2, Krishna S Nayak1.   

Abstract

PURPOSE: To apply deep convolution neural network to the segmentation task in myocardial arterial spin labeled perfusion imaging and to develop methods that measure uncertainty and that adapt the convolution neural network model to a specific false-positive versus false-negative tradeoff.
METHODS: The Monte Carlo dropout U-Net was trained on data from 22 subjects and tested on data from 6 heart transplant recipients. Manual segmentation and regional myocardial blood flow were available for comparison. We consider 2 global uncertainty measures, named "Dice uncertainty" and "Monte Carlo dropout uncertainty," which were calculated with and without the use of manual segmentation, respectively. Tversky loss function with a hyperparameter β was used to adapt the model to a specific false-positive versus false-negative tradeoff.
RESULTS: The Monte Carlo dropout U-Net achieved a Dice coefficient of 0.91 ± 0.04 on the test set. Myocardial blood flow measured using automatic segmentations was highly correlated to that measured using the manual segmentation (R2 = 0.96). Dice uncertainty and Monte Carlo dropout uncertainty were in good agreement (R2 = 0.64). As β increased, the false-positive rate systematically decreased and false-negative rate systematically increased.
CONCLUSION: We demonstrate the feasibility of deep convolution neural network for automatic segmentation of myocardial arterial spin labeling, with good accuracy. We also introduce 2 simple methods for assessing model uncertainty. Finally, we demonstrate the ability to adapt the convolution neural network model to a specific false-positive versus false-negative tradeoff. These findings are directly relevant to automatic segmentation in quantitative cardiac MRI and are broadly applicable to automatic segmentation problems in diagnostic imaging.
© 2019 International Society for Magnetic Resonance in Medicine.

Entities:  

Keywords:  Bayesian; Monte Carlo dropout; automatic segmentation; deep convolutional neural network; false-positive and false-negative tradeoff; uncertainty measure

Mesh:

Year:  2019        PMID: 31729078      PMCID: PMC6989045          DOI: 10.1002/mrm.28043

Source DB:  PubMed          Journal:  Magn Reson Med        ISSN: 0740-3194            Impact factor:   4.668


  28 in total

1.  SENSE: sensitivity encoding for fast MRI.

Authors:  K P Pruessmann; M Weiger; M B Scheidegger; P Boesiger
Journal:  Magn Reson Med       Date:  1999-11       Impact factor: 4.668

2.  Quantification of relative cerebral blood flow change by flow-sensitive alternating inversion recovery (FAIR) technique: application to functional mapping.

Authors:  S G Kim
Journal:  Magn Reson Med       Date:  1995-09       Impact factor: 4.668

3.  Fully automatic segmentation of the brain in MRI.

Authors:  M S Atkins; B T Mackiewich
Journal:  IEEE Trans Med Imaging       Date:  1998-02       Impact factor: 10.048

4.  Deep convolutional neural network and 3D deformable approach for tissue segmentation in musculoskeletal magnetic resonance imaging.

Authors:  Fang Liu; Zhaoye Zhou; Hyungseok Jang; Alexey Samsonov; Gengyan Zhao; Richard Kijowski
Journal:  Magn Reson Med       Date:  2017-07-21       Impact factor: 4.668

5.  Double-gated myocardial ASL perfusion imaging is robust to heart rate variation.

Authors:  Hung Phi Do; Andrew J Yoon; Michael W Fong; Farhood Saremi; Mark L Barr; Krishna S Nayak
Journal:  Magn Reson Med       Date:  2016-05-30       Impact factor: 4.668

6.  MR perfusion studies with T1-weighted echo planar imaging.

Authors:  K K Kwong; D A Chesler; R M Weisskoff; K M Donahue; T L Davis; L Ostergaard; T A Campbell; B R Rosen
Journal:  Magn Reson Med       Date:  1995-12       Impact factor: 4.668

7.  Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation.

Authors:  Konstantinos Kamnitsas; Christian Ledig; Virginia F J Newcombe; Joanna P Simpson; Andrew D Kane; David K Menon; Daniel Rueckert; Ben Glocker
Journal:  Med Image Anal       Date:  2016-10-29       Impact factor: 8.545

8.  Automated Extracellular Volume Fraction Mapping Provides Insights Into the Pathophysiology of Left Ventricular Remodeling Post-Reperfused ST-Elevation Myocardial Infarction.

Authors:  Heerajnarain Bulluck; Stefania Rosmini; Amna Abdel-Gadir; Steven K White; Anish N Bhuva; Thomas A Treibel; Marianna Fontana; Esther Gonzalez-Lopez; Patricia Reant; Manish Ramlall; Ashraf Hamarneh; Alex Sirker; Anna S Herrey; Charlotte Manisty; Derek M Yellon; Peter Kellman; James C Moon; Derek J Hausenloy
Journal:  J Am Heart Assoc       Date:  2016-07-11       Impact factor: 5.501

9.  Extracellular volume fraction mapping in the myocardium, part 1: evaluation of an automated method.

Authors:  Peter Kellman; Joel R Wilson; Hui Xue; Martin Ugander; Andrew E Arai
Journal:  J Cardiovasc Magn Reson       Date:  2012-09-10       Impact factor: 5.364

10.  Myocardial arterial spin labeling perfusion imaging with improved sensitivity.

Authors:  Hung Phi Do; Terrence R Jao; Krishna S Nayak
Journal:  J Cardiovasc Magn Reson       Date:  2014-01-27       Impact factor: 5.364

View more
  1 in total

1.  Deep learning applications in myocardial perfusion imaging, a systematic review and meta-analysis.

Authors:  Ebraham Alskaf; Utkarsh Dutta; Cian M Scannell; Amedeo Chiribiri
Journal:  Inform Med Unlocked       Date:  2022
  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.