Background Despite advances, blood oxygen level-dependent (BOLD) cardiac MRI for myocardial perfusion is limited by inadequate spatial coverage, imaging speed, multiple breath holds, and imaging artifacts, particularly at 3.0 T. Purpose To develop and validate a robust, contrast agent-unenhanced, free-breathing three-dimensional (3D) cardiac MRI approach for reliably examining changes in myocardial perfusion between rest and adenosine stress. Materials and Methods A heart rate-independent, free-breathing 3D T2 mapping technique at 3.0 T that can be completed within the period of adenosine stress (≤4 minutes) was developed by using computer simulations, ex vivo heart preparations, and dogs. Studies in dogs were performed with and without coronary stenosis and validated with simultaneously acquired nitrogen 13 (¹³N) ammonia PET perfusion in a clinical PET/MRI system. The MRI approach was also prospectively evaluated in healthy human volunteers (from January 2017 to September 2017). Myocardial BOLD responses (MBRs) between normal and ischemic myocardium were compared with mixed model analysis. Results Dogs (n = 10; weight range, 20-25 kg; mongrel dogs) and healthy human volunteers (n = 10; age range, 22-53 years; seven men) were evaluated. In healthy dogs, T2 MRI at adenosine stress was greater than at rest (mean rest vs stress, 38.7 msec ± 2.5 [standard deviation] vs 45.4 msec ± 3.3, respectively; MBR, 1.19 ± 0.08; both, P < .001). At the same conditions, mean rest versus stress PET perfusion was 1.1 mL/mg/min ± 0.11 versus 2.3 mL/mg/min ± 0.82, respectively (P < .001); myocardial perfusion reserve (MPR) was 2.4 ± 0.82 (P < .001). The BOLD response and PET MPR were positively correlated (R = 0.67; P < .001). In dogs with coronary stenosis, perfusion anomalies were detected on the basis of MBR (normal vs ischemic, 1.09 ± 0.05 vs 1.00 ± 0.04, respectively; P < .001) and MPR (normal vs ischemic, 2.7 ± 0.08 vs 1.7 ± 1.1, respectively; P < .001). Human volunteers showed increased myocardial T2 at stress (rest vs stress, 44.5 msec ± 2.6 vs 49.0 msec ± 5.5, respectively; P = .004; MBR, 1.1 msec ± 8.08). Conclusion This three-dimensional cardiac blood oxygen level-dependent (BOLD) MRI approach overcame key limitations associated with conventional cardiac BOLD MRI by enabling whole-heart coverage within the standard duration of adenosine infusion, and increased the magnitude and reliability of BOLD contrast, which may be performed without requiring breath holds. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Almeida in this issue.

Patient motion degrades image quality, affecting the quantitative assessment of PET images. This problem affects studies of coronary lesions in which microcalcification processes are targeted. Coronary PET imaging protocols require scans of up to 30 min, introducing the risk of gross patient motion (GPM) during the acquisition. Here, we investigate the feasibility of an automated data-driven method for the detection of GPM during PET acquisition. Methods: Twenty-eight patients with stable coronary disease underwent a 30-min PET acquisition 1 h after the injection of ¹⁸F-sodium fluoride (¹⁸F-NaF) at 248 ± 10 MBq (mean ± SD) and then a coronary CT angiography scan. An automated data-driven GPM detection technique tracking the center of mass of the count rates for every 200 ms in the PET list-mode data was devised and evaluated. Two patient motion patterns were considered: sudden repositioning (motion of >0.5 mm within 3 s) and general repositioning (motion of >0.3 mm over 15 s or more). After the reconstruction of diastolic images, individual GPM frames with focal coronary uptake were coregistered in 3 dimensions, creating a GPM-compensated (GPMC) image series. Lesion motion was reported for all lesions with focal uptake. Relative differences in SUV_max and target-to-background ratio (TBR) between GPMC and non-GPMC (standard electrocardiogram-gated data) diastolic PET images were compared in 3 separate groups defined by the maximum motion observed in the lesion (<5, 5-10, and >10 mm). Results: A total of 35 ¹⁸F-NaF-avid lesions were identified in 28 patients. An average of 3.5 ± 1.5 GPM frames were considered for each patient, resulting in an average frame duration of 7 ± 4 (range, 3-21) min. The mean per-patient motion was: 7 ± 3 mm (maximum, 13.7 mm). GPM correction increased SUV_max and TBR in all lesions with greater than 5 mm of motion. In lesions with 5-10 mm of motion (n = 15), SUV_max and TBR increased by 4.6% ± 5.6% (P = 0.02) and 5.8% ± 6.4% (P < 0.002), respectively. In lesions with greater than 10 mm of motion (n = 15), the SUV_max and TBR increased by 5.0% ± 5.3% (P = 0.009) and 11.5% ± 10.1% (P = 0.001), respectively. GPM correction led to the diagnostic reclassification of 3 patients (11%). Conclusion: GPM during coronary ¹⁸F-NaF PET imaging is common and may affect quantitative accuracy. Automated retrospective compensation of this motion is feasible and should be considered for coronary PET imaging.

Background

Challenges to cardiac PET-CT include patient motion, prolonged image acquisition and a reduction of counts due to gating. We compared two analytical tools, FusionQuant and OsiriX, for quantification of gated cardiac 18F-sodium fluoride (18F-fluoride) PET-CT imaging.

Methods

Twenty-seven patients with aortic stenosis were included, 15 of whom underwent repeated imaging 4 weeks apart. Agreement between analytical tools and scan-rescan reproducibility was determined using the Bland-Altman method and Lin's concordance correlation coefficients (CCC).

Results

Image analysis was faster with FusionQuant [median time (IQR) 7:10 (6:40-8:20) minutes] compared with OsiriX [8:30 (8:00-10:10) minutes, p = .002]. Agreement of uptake measurements between programs was excellent, CCC = 0.972 (95% CI 0.949-0.995) for mean tissue-to-background ratio (TBR_mean) and 0.981 (95% CI 0.965-0.997) for maximum tissue-to-background ratio (TBR_max). Mean noise decreased from 11.7% in the diastolic gate to 6.7% in motion-corrected images (p = .002); SNR increased from 25.41 to 41.13 (p = .0001). Aortic valve scan-rescan reproducibility for TBR_max was improved with FusionQuant using motion correction compared to OsiriX (error ± 36% vs ± 13%, p < .001) while reproducibility for TBR_mean was similar (± 10% vs ± 8% p = .252).

Conclusion

18F-fluoride PET quantification with FusionQuant and OsiriX is comparable. FusionQuant with motion correction offers advantages with respect to analysis time and reproducibility of TBR_max values.

Background

PET reconstruction incorporating spatially variant 3D Point Spread Function (PSF) improves contrast and image resolution. "Cardiac Motion Frozen" (CMF) processing eliminates the influence of cardiac motion in static summed images. We have evaluated the combined use of CMF- and PSF-based reconstruction for high-resolution cardiac PET.

Methods

Static and 16-bin ECG-gated images of 20 patients referred for (18)F-FDG myocardial viability scans were obtained on a Siemens Biograph-64. CMF was applied to the gated images reconstructed with PSF. Myocardium to blood contrast, maximum left ventricle (LV) counts to defect contrast, contrast-to-noise (CNR) and wall thickness with standard reconstruction (2D-AWOSEM), PSF, ED-gated PSF, and CMF-PSF were compared.

Results

The measured wall thickness was 18.9 ± 5.2 mm for 2D-AWOSEM, 16.6 ± 4.5 mm for PSF, and 13.8 ± 3.9 mm for CMF-PSF reconstructed images (all P < .05). The CMF-PSF myocardium to blood and maximum LV counts to defect contrasts (5.7 ± 2.7, 10.0 ± 5.7) were higher than for 2D-AWOSEM (3.5 ± 1.4, 6.5 ± 3.1) and for PSF (3.9 ± 1.7, 7.7 ± 3.7) (CMF vs all other, P < .05). The CNR for CMF-PSF (26.3 ± 17.5) was comparable to PSF (29.1 ± 18.3), but higher than for ED-gated dataset (13.7 ± 8.8, P < .05).

Conclusion

Combined CMF-PSF reconstruction increased myocardium to blood contrast, maximum LV counts to defect contrast and maintained equivalent noise when compared to static summed 2D-AWOSEM and PSF reconstruction.

Background

Coronary artery calcification (CAC), often assessed by computed tomography (CT), is a powerful marker of coronary artery disease that can guide preventive therapies. Computed tomographies, however, are not always accessible or serially obtainable. It remains unclear whether other widespread tests such as transthoracic echocardiograms (TTEs) can be used to predict CAC.

Methods

Using a data set of 2,881 TTE videos paired with coronary calcium CTs, we trained a video-based artificial intelligence convolutional neural network to predict CAC scores from parasternal long-axis views. We evaluated the model's ability to classify patients from a held-out sample as well as an external site sample into zero CAC and high CAC (CAC ≥ 400 Agatston units) groups by receiver operating characteristic and precision-recall curves. We also investigated whether such classifications prognosticated significant differences in 1-year mortality rates by the log-rank test of Kaplan-Meier curves.

Results

Transthoracic echocardiogram artificial intelligence models had high discriminatory abilities in predicting zero CAC (receiver operating characteristic area under the curve [AUC] = 0.81 [95% CI, 0.74-0.88], F1 score = 0.95) and high CAC (AUC = 0.74 [0.68-0.8], F1 score = 0.74). This performance was confirmed in an external test data set of 92 TTEs (AUC = 0.75 [0.65-0.85], F1 score = 0.77; and AUC = 0.85 [0.76-0.93], F1 score = 0.59, respectively). Risk stratification by TTE-predicted CAC performed similarly to CT CAC scores in prognosticating significant differences in 1-year survival in high-CAC patients (CT CAC ≥ 400 vs CT CAC < 400, P = .03; TTE-predicted CAC ≥ 400 vs TTE-predicted CAC < 400, P = .02).

Conclusions

A video-based deep learning model successfully used TTE videos to predict zero CAC and high CAC with high accuracy. Transthoracic echocardiography-predicted CAC prognosticated differences in 1-year survival similar to CT CAC. Deep learning of TTEs holds promise for future adjunctive coronary artery disease risk stratification to guide preventive therapies.

Objective

We aimed to investigate if early revascularization in patients with suspected coronary artery disease can be effectively predicted by integrating clinical data and quantitative image features derived from perfusion SPECT (MPS) by machine learning (ML) approach.

Methods

713 rest (201)Thallium/stress (99m)Technetium MPS studies with correlating invasive angiography with 372 revascularization events (275 PCI/97 CABG) within 90 days after MPS (91% within 30 days) were considered. Transient ischemic dilation, stress combined supine/prone total perfusion deficit (TPD), supine rest and stress TPD, exercise ejection fraction, and end-systolic volume, along with clinical parameters including patient gender, history of hypertension and diabetes mellitus, ST-depression on baseline ECG, ECG and clinical response during stress, and post-ECG probability by boosted ensemble ML algorithm (LogitBoost) to predict revascularization events. These features were selected using an automated feature selection algorithm from all available clinical and quantitative data (33 parameters). Tenfold cross-validation was utilized to train and test the prediction model. The prediction of revascularization by ML algorithm was compared to standalone measures of perfusion and visual analysis by two experienced readers utilizing all imaging, quantitative, and clinical data.

Results

The sensitivity of machine learning (ML) (73.6% ± 4.3%) for prediction of revascularization was similar to one reader (73.9% ± 4.6%) and standalone measures of perfusion (75.5% ± 4.5%). The specificity of ML (74.7% ± 4.2%) was also better than both expert readers (67.2% ± 4.9% and 66.0% ± 5.0%, P < .05), but was similar to ischemic TPD (68.3% ± 4.9%, P < .05). The receiver operator characteristics areas under curve for ML (0.81 ± 0.02) was similar to reader 1 (0.81 ± 0.02) but superior to reader 2 (0.72 ± 0.02, P < .01) and standalone measure of perfusion (0.77 ± 0.02, P < .01).

Conclusion

ML approach is comparable or better than experienced readers in prediction of the early revascularization after MPS, and is significantly better than standalone measures of perfusion derived from MPS.

Purpose

¹⁸F-sodium fluoride (¹⁸F-NaF) has shown promise in assessing disease activity in coronary arteries, but currently used measures of activity - such as maximum target to background ratio (TBRmax) - are defined by single pixel count values. We aimed to develop a novel coronary-specific measure of ¹⁸F-NaF PET reflecting activity throughout the entire coronary vasculature (coronary microcalcification activity [CMA]).

Methods

Patients with recent myocardial infarction and multi-vessel coronary artery disease underwent ¹⁸F-NaF PET and coronary CT angiography. We assessed the association between coronary ¹⁸F-NaF uptake (both TBRmax and CMA) and coronary artery calcium scores (CACS) as well as low attenuation plaque (LAP, attenuation < 30 Hounsfield units) volume.

Results

In 50 patients (64% males, 63 ± 7 years), CMA and TBRmax were higher in vessels with LAP compared to those without LAP (1.09 [0.02, 2.34] versus 0.0 [0.0, 0.0], p < 0.001 and 1.23 [1.16, 1.37] versus 1.04 [0.93, 1.11], p < 0.001). Compared to a TBRmax threshold of 1.25, CMA > 0 had a higher diagnostic accuracy for detection of LAP: sensitivity of 93.1 (83.3-98.1)% versus 58.6 (44.9-71.4)% and a specificity of 95.7 (88.0-99.1)% versus 80.0 (68.7-88.6)% (both p < 0.001). ¹⁸F-NaF uptake assessed by CMA correlated more closely with LAP (r = 0.86, p < 0.001) than the CT calcium score (r = 0.39, p < 0.001), with these associations outperforming those observed for TBRmax values (LAP r = 0.63, p < 0.001; CT calcium score r = 0.30, p < 0.001).

Conclusions

Automated assessment of disease activity across the entire coronary vasculature is feasible using ¹⁸F-NaF CMA, providing a single measurement that has closer agreement with CT markers of plaque vulnerability than more traditional measures of plaque activity.

Background

Current ¹⁸F-NaF assessments of aortic valve microcalcification using ¹⁸F-NaF PET/CT are based on evaluations of end-diastolic or cardiac motion-corrected (ECG-MC) images, which are affected by both patient and respiratory motion. We aimed to test the impact of employing a triple motion correction technique (3 × MC), including cardiorespiratory and gross patient motion, on quantitative and qualitative measurements.

Materials and methods

Fourteen patients with aortic stenosis underwent two repeat 30-min PET aortic valve scans within (29 ± 24) days. We considered three different image reconstruction protocols; an end-diastolic reconstruction protocol (standard) utilizing 25% of the acquired data, an ECG-gated (four ECG gates) reconstruction (ECG-MC), and a triple motion-corrected (3 × MC) dataset which corrects for both cardiorespiratory and patient motion. All datasets were compared to aortic valve calcification scores (AVCS), using the Agatston method, obtained from CT scans using correlation plots. We report SUV_max values measured in the aortic valve and maximum target-to-background ratios (TBR_max) values after correcting for blood pool activity.

Results

Compared to standard and ECG-MC reconstructions, increases in both SUV_max and TBR_max were observed following 3 × MC (SUV_max: Standard = 2.8 ± 0.7, ECG-MC = 2.6 ± 0.6, and 3 × MC = 3.3 ± 0.9; TBR_max: Standard = 2.7 ± 0.7, ECG-MC = 2.5 ± 0.6, and 3 × MC = 3.3 ± 1.2, all p values ≤ 0.05). 3 × MC had improved correlations (R² value) to the AVCS when compared to the standard methods (SUV_max: Standard = 0.10, ECG-MC = 0.10, and 3 × MC = 0.20; TBR_max: Standard = 0.20, ECG-MC = 0.28, and 3 × MC = 0.46).

Conclusion

3 × MC improves the correlation between the AVCS and SUV_max and TBR_max and should be considered in PET studies of aortic valves using ¹⁸F-NaF.

Background

We investigated the relationship of quantitative plaque features from coronary computed tomography (CT) angiography and coronary vascular dysfunction by impaired myocardial flow reserve (MFR) by (13)N-Ammonia positron emission tomography (PET).

Methods and results

Fifty-one patients (32 men, 62.4±9.5 years) underwent combined rest-stress (13)N-ammonia PET and CT angiography scans by hybrid PET/CT. Regional MFR was measured from PET. From CT angiography, 153 arteries were evaluated by semiautomated software, computing arterial noncalcified plaque (NCP), low-density NCP (NCP<30 HU), calcified and total plaque volumes, and corresponding plaque burden (plaque volumex100%/vessel volume), stenosis, remodeling index, contrast density difference (maximum difference in luminal attenuation per unit area in the lesion), and plaque length. Quantitative stenosis, plaque burden, and myocardial mass were combined by boosted ensemble machine-learning algorithm into a composite risk score to predict impaired MFR (MFR≤2.0) by PET in each artery. Nineteen patients had impaired regional MFR in at least 1 territory (41/153 vessels). Patients with impaired regional MFR had higher arterial NCP (32.4% versus 17.2%), low-density NCP (7% versus 4%), and total plaque burden (37% versus 19.3%, P<0.02). In multivariable analysis with 10-fold cross-validation, NCP burden was the most significant predictor of impaired MFR (odds ratio, 1.35; P=0.021 for all). For prediction of impaired MFR with 10-fold cross-validation, receiver operating characteristics area under the curve for the composite score was 0.83 (95% confidence interval, 0.79-0.91) greater than for quantitative stenosis (0.66, 95% confidence interval, 0.57-0.76, P=0.005).

Conclusions

Compared with stenosis, arterial NCP burden and a composite score combining quantitative stenosis and plaque burden from CT angiography significantly improves identification of downstream regional vascular dysfunction.

Purpose

To improve the test-retest reproducibility of coronary plaque ¹⁸F-sodium fluoride (¹⁸F-NaF) positron emission tomography (PET) uptake measurements.

Methods

We recruited 20 patients with coronary artery disease who underwent repeated hybrid PET/CT angiography (CTA) imaging within 3 weeks. All patients had 30-min PET acquisition and CTA during a single imaging session. Five PET image-sets with progressive motion correction were reconstructed: (i) a static dataset (no-MC), (ii) end-diastolic PET (standard), (iii) cardiac motion corrected (MC), (iv) combined cardiac and gross patient motion corrected (2 × MC) and, (v) cardiorespiratory and gross patient motion corrected (3 × MC). In addition to motion correction, all datasets were corrected for variations in the background activities which are introduced by variations in the injection-to-scan delays (background blood pool clearance correction, BC). Test-retest reproducibility of PET target-to-background ratio (TBR) was assessed by Bland-Altman analysis and coefficient of reproducibility.

Results

A total of 47 unique coronary lesions were identified on CTA. Motion correction in combination with BC improved the PET TBR test-retest reproducibility for all lesions (coefficient of reproducibility: standard = 0.437, no-MC = 0.345 (27% improvement), standard + BC = 0.365 (20% improvement), no-MC + BC = 0.341 (27% improvement), MC + BC = 0.288 (52% improvement), 2 × MC + BC = 0.278 (57% improvement) and 3 × C + BC = 0.254 (72% improvement), all p < 0.001). Importantly, in a sub-analysis of ¹⁸F-NaF-avid lesions with gross patient motion > 10 mm following corrections, reproducibility was improved by 133% (coefficient of reproducibility: standard = 0.745, 3 × MC = 0.320).

Conclusion

Joint corrections for cardiac, respiratory, and gross patient motion in combination with background blood pool corrections markedly improve test-retest reproducibility of coronary ¹⁸F-NaF PET.