Assessment of segmental myocardial blood flow and myocardial perfusion reserve by adenosine‐stress myocardial arterial spin labeling perfusion imaging

Yoon, Andrew; Do, Hung P; Cen, Steven; Fong, Michael W.; Saremi, Farhood; Barr, Mark L.; Nayak, Krishna S.

doi:10.1002/jmri.25604

Cited by 11 publications

(17 citation statements)

References 50 publications

(102 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Solid line is the 50 C isotherm, which can be assumed to be the thermal lesion contour. Dashed line represents the location of this contour when the blood perfusion term is included at a rate of 1719 ml/min/kg (maximum value reported in [17], but very similar to that measured in resting humans (1750 ± 860 ml/min/kg) using perfusion imaging techniques [19]).…”

Section: Resultssupporting

confidence: 54%

“…The increase in the difference over time is clearly noticeable. This case is relevant, since a recent study reported an even higher perfusion value for resting humans (1750 ± 860 ml/min/kg [19]). Figure 3 shows the progress of applied power (between 6 and 7 W) in the case of constant power ablation, and electrode temperature (50 C target) in the case of constant temperature ablation.…”

Section: Resultsmentioning

confidence: 90%

“…Consequently, the effect of the intramyocardial blood perfusion on lesion size under constant power and an irrigated electrode are still unknown. Our goal was thus to assess the impact of its inclusion in RF cardiac ablation computer models in terms of thermal lesion depth and to consider how information obtained from perfusion imaging techniques (e.g., those based on magnetic resonance imaging (MRI) [19]) could be employed to tune computer-based predictive tools.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Numerical analysis of thermal impact of intramyocardial capillary blood flow during radiofrequency cardiac ablation

Pérez

González-Suárez

Berjano

2017

International Journal of Hyperthermia

View full text Add to dashboard Cite

show abstract

Section: Resultssupporting

confidence: 54%

Section: Resultsmentioning

confidence: 90%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Numerical analysis of thermal impact of intramyocardial capillary blood flow during radiofrequency cardiac ablation

Pérez

González-Suárez

Berjano

2017

International Journal of Hyperthermia

View full text Add to dashboard Cite

show abstract

“…Previous attempts at myocardial ASL have employed a minimum TR of approximately six seconds to allow for a complete T1 recovery of the label (1664 ms at 3 T) . This long TR decreases scan efficiency and limits the total number of pairs to be averaged in a reasonable acquisition time.…”

Section: Introductionmentioning

confidence: 99%

Optimal repetition time for free breathing myocardial arterial spin labeling

et al. 2019

View full text Add to dashboard Cite

Arterial spin labeling (ASL) offers a noninvasive and repeatable measurement of tissue blood flow, but its application in the human heart is still challenging. The aim of this study was to improve the scan efficiency of flow-sensitive alternating inversion recovery (FAIR) ASL using a single inversion time. First, by evaluating a free breathing acquisition without navigators. Second, by shortening TR to acquire more ASL pairs.A FAIR ASL with balanced steady state free precession (bSSFP) readout was implemented at 3 T and tested in 14 healthy volunteers. Three cardiac-triggered ASL scans with different TRs were compared while maintaining a constant acquisition time.These scans were named by the trigger pulse (TP) that started each acquisition: every two (TP1), four (TP2) and six (TP3) cardiac cycles. Two studies were performed on different days to assess the reproducibility of TP2 and TP3 scans. Perfusion response during passive leg raising was also evaluated. Student t-test and nonparametric equivalent were computed for perfusion differences. Bland-Altman and coefficient of variation (CV) were calculated for the reproducibility assessment.Mean myocardial blood flow (MBF) values measured at rest were: 1.47 ± 0.91 ml/g/min (TP1), 1.95 ± 0.93 ml/g/min (TP2) and 1.94 ± 1.41 ml/g/min (TP3). Intrasession CV was 17.15% (TP2) and 24.17% (TP3) and intersession CV was 36.86% (TP2) and 18.38% (TP3). Mean MBF was found to increase with passive stress by 0.22 ± 0.41 ml/g/min for TP2 (P = 0.037).Mean MBF values were consistent with the literature, although in the high end of the normal range. The lower MBF values measured in TP1 scans are attributable to saturation of blood outside the slice because of shorter TR. Free breathing FAIR cardiac ASL without navigators is feasible. A TR of approximately four seconds (TP2) offers the best alternative with similar mean ASL and tSNR than longer TR acquisitions.

show abstract

“…Trained networks were tested on 144 unseen ASL images acquired at rest and during adenosine stress from 6 heart transplant recipients. Training, validation, and testing data were drawn from previously published studies (31)(32)(33), where manual segmentation and quantitative reginal MBF from all subjects were readily available.…”

Section: Network Architecturementioning

confidence: 99%

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

Guo

Yoon

et al. 2019

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

words maximum)Purpose: To apply deep CNN to the segmentation task in myocardial arterial spin labeled (ASL) perfusion imaging and to develop methods that measure uncertainty and that adapt the CNN model to a specific false positive vs. false negative tradeoff. Methods:The Monte Carlo dropout (MCD) 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 (MBF) were available for comparison. We consider two global uncertainty measures, named "Dice Uncertainty" and "MCD 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 vs. false negative tradeoff. Results:The MCD U-Net achieved Dice coefficient of 0.91 ± 0.04 on the test set. MBF measured using automatic segmentations was highly correlated to that measured using the manual segmentation (R 2 = 0.96). Dice Uncertainty and MCD Uncertainty were in good agreement (R 2 = 0.64). As β increased, the false positive rate systematically decreased and false negative rate systematically increased. Conclusion:We demonstrate the feasibility of deep CNN for automatic segmentation of myocardial ASL, with good accuracy. We also introduce two simple methods for assessing model uncertainty. Finally, we demonstrate the ability to adapt the CNN model to a specific false positive vs. 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.Keywords: MRI, arterial spin labeling, automatic segmentation, deep convolutional neural network, false positive and false negative tradeoff, uncertainty measure, quality assessment, Bayesian, Monte Carlo Dropout

show abstract

Assessment of segmental myocardial blood flow and myocardial perfusion reserve by adenosine‐stress myocardial arterial spin labeling perfusion imaging

Abstract: 2 Technical Efficacy: Stage 1 J. MAGN. RESON. IMAGING 2017;46:413-420.

Cited by 11 publications

References 50 publications

Numerical analysis of thermal impact of intramyocardial capillary blood flow during radiofrequency cardiac ablation

Numerical analysis of thermal impact of intramyocardial capillary blood flow during radiofrequency cardiac ablation

Optimal repetition time for free breathing myocardial arterial spin labeling

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

Contact Info

Product

Resources

About