Magnetic resonance perfusion imaging in ischemic heart disease

Deep learning-based segmentation of left ventricular myocardium on dynamic contrast-enhanced MRI: a comprehensive evaluation across temporal frames

PURPOSE

Cardiac perfusion MRI is vital for disease diagnosis, treatment planning, and risk stratification, with anomalies serving as markers of underlying ischemic pathologies. AI-assisted methods and tools enable accurate and efficient left ventricular (LV) myocardium segmentation on all DCE-MRI timeframes, offering a solution to the challenges posed by the multidimensional nature of the data. This study aims to develop and assess an automated method for LV myocardial segmentation on DCE-MRI data of a local hospital.

METHODS

The study consists of retrospective DCE-MRI data from 55 subjects acquired at the local hospital using a 1.5 T MRI scanner. The dataset included subjects with and without cardiac abnormalities. The timepoint for the reference frame (post-contrast LV myocardium) was identified using standard deviation across the temporal sequences. Iterative image registration of other temporal images with respect to this reference image was performed using Maxwell's demons algorithm. The registered stack was fed to the model built using the U-Net framework for predicting the LV myocardium at all timeframes of DCE-MRI.

RESULTS

The mean and standard deviation of the dice similarity coefficient (DSC) for myocardial segmentation using pre-trained network Net_cine is 0.78 ± 0.04, and for the fine-tuned network Net_dyn which predicts mask on all timeframes individually, it is 0.78 ± 0.03. The DSC for Net_dyn ranged from 0.71 to 0.93. The average DSC achieved for the reference frame is 0.82 ± 0.06.

CONCLUSION

The study proposed a fast and fully automated AI-assisted method to segment LV myocardium on all timeframes of DCE-MRI data. The method is robust, and its performance is independent of the intra-temporal sequence registration and can easily accommodate timeframes with potential registration errors.

Assessment of Myocardial Perfusion at Rest and During Stress Using Dynamic First-Pass Contrast-Enhanced Magnetic Resonance Imaging in Healthy Dogs

Objective: To assess the feasibility of myocardial perfusion analysis in healthy dogs using dynamic contrast-enhanced cardiac magnetic resonance (DCE-MR) imaging at rest and during simulated stress with two doses of adenosine.

Animals: Ten healthy beagle dogs.

Procedures: Dogs were anesthetized and positioned in dorsal recumbency in a 3.0 Tesla MR scanner. Electrocardiogram-triggered dynamic T1-weighted ultrafast gradient echo images of three slices in short-axis orientation of the heart were acquired during breath holds and the first pass of gadolinium contrast. Image acquisition was performed after 4 min infusion of 140 μg/kg/min and 280 μg/kg/min adenosine and, after a washout period, without adenosine, respectively. Images were processed by dividing each slice into 6 radial segments and perfusion analysis was performed from signal intensity-time data.

Results: No differences in perfusion parameters were found between segments within any of the slices, but significant differences were found between slices for peak enhancement, accumulated enhancement, and the maximum upslope. In addition, significant differences were found within each slice between data at rest and during adenosine-induced stress for the relative and absolute maximum upslope, relative peak enhancement, time to peak, and accumulated enhancement although inter-individual variation was large and no difference was found between the two stress tests for some parameters.

Conclusion and Clinical Relevance: Results of this study showed that rest and stress myocardial perfusion can be assessed using DCE-CMR in dogs using the methods described. Both, adenosine dose and slice appear to affect perfusion parameters in healthy dogs and individual response to adenosine was variable.

Cost-effectiveness of initial stress cardiovascular MR, stress SPECT or stress echocardiography as a gate-keeper test, compared with upfront invasive coronary angiography in the investigation and management of patients with stable chest pain: mid-term outcomes from the CECaT randomised controlled trial

OBJECTIVES

To compare outcomes and cost-effectiveness of various initial imaging strategies in the management of stable chest pain in a long-term prospective randomised trial.

SETTING

Regional cardiothoracic referral centre in the east of England.

PARTICIPANTS

898 patients (69% man) entered the study with 869 alive at 2 years of follow-up. Patients were included if they presented for assessment of stable chest pain with a positive exercise test and no prior history of ischaemic heart disease. Exclusion criteria were recent infarction, unstable symptoms or any contraindication to stress MRI.

PRIMARY OUTCOME MEASURES

The primary outcomes of this follow-up study were survival up to a minimum of 2 years post-treatment, quality-adjusted survival and cost-utility of each strategy.

RESULTS

898 patients were randomised. Compared with angiography, mortality was marginally higher in the groups randomised to cardiac MR (HR 2.6, 95% CI 1.1 to 6.2), but similar in the single photon emission CT-methoxyisobutylisonitrile (SPECT-MIBI; HR 1.0, 95% CI 0.4 to 2.9) and ECHO groups (HR 1.6, 95% CI 0.6 to 4.0). Although SPECT-MIBI was marginally superior to other non-invasive tests there were no other significant differences between the groups in mortality, quality-adjusted survival or costs.

CONCLUSIONS

Non-invasive cardiac imaging can be used safely as the initial diagnostic test to diagnose coronary artery disease without adverse effects on patient outcomes or increased costs, relative to angiography. These results should be interpreted in the context of recent advances in imaging technology.

TRIAL REGISTRATION

ISRCTN 47108462, UKCRN 3696.

Evaluation of rigid and non-rigid motion compensation of cardiac perfusion MRI

Although the evaluation of cardiac perfusion using MRI could be of crucial importance for the diagnosis of ischemic heart diseases, it is still not a routinely used technique. The major difficulty is that MR perfusion images are often corrupted by inconsistent myocardial motion. Although motion compensation methods have been studied throughout the past decade, no clinically accepted solution has emerged. This is partly due to the lack of comprehensive validation. To address this deficit we collected a large multi-centre MR perfusion dataset and used this to characterize typical myocardial motion and confirmed that under clinically relevant conditions motion correction is a frequent requirement (67% of all 586 cases). We then developed a proposed solution which includes both rigid/affine and the non-rigid image registration. Quantitative validation has been conducted using 6 different statistics to provide a comprehensive evaluation, showing the proposed techniques to be highly robust to different myocardial anatomy and motion patterns as well as to MR imaging acquisition parameters.

Usage of the T1 effect of an iron oxide contrast agent in an animal model to quantify myocardial blood flow by MRI

BACKGROUND

To present a new method for fully quantitative analysis of myocardial blood flow (MBF) using magnetic resonance imaging. The first pass of an intravascular iron oxide contrast medium can be used to quantify myocardial perfusion. The technique was validated in an animal model using colored microspheres.

MATERIALS AND METHODS

In six pigs, a tracking catheter was positioned in the left anterior descending artery (LAD). Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) was performed on a 1.5-T scanner using a hybrid gradient-echo/echoplanar imaging (GRE-EPI) sequence. Regional myocardial blood flow (rMBF) was altered by either inducing vasodilatation with adenosine or creating coronary artery obstruction. The T(1) effect of a superparamagnetic iron oxide-based contrast medium (Resovist) administered at a dose of 8 micromol/kg was used. Upslope, time-to-peak and peak intensity were calculated from the signal intensity-time curves and absolute rMBF using the Kety-Schmidt equation; results were compared to those obtained using colored microspheres.

RESULTS

The mean rMBF calculated by MRI was 1.49 (+/-6.91, quartile width) ml/min/g versus 3.21 (+/-1.61) ml/min/g measured by means of microspheres under resting conditions. rMBF increased to a mean of 6.21 (+/-2.83) ml/min/g versus 4.22 (+/-1.70) ml/min/g under adenosine and was reduced to zero flow in total occlusion. Linear regression showed the best correlation for upslope (R=0.714), time-to-peak (R=0.626) and the Kety-Schmidt equation (R=0.584).

CONCLUSIONS

The T(1) effect of an iron oxide-based contrast medium allows determination of rMBF when using the Kety-Schmidt equation. The results are similar to those obtained with the standard of reference, colored micropheres, but not better than the results of the semiquantitative approach.

In vivo R1-enhancement mapping of canine myocardium using ceMRI with Gd(ABE-DTTA) in an acute ischemia-reperfusion model

PURPOSE

To demonstrate the usefulness of normalized DeltaR1 (DeltaR1(n)) mapping in myocardial tissue following the administration of the contrast agent (CA) Gd(ABE-DTTA).

MATERIALS AND METHODS

Ischemia-reperfusion experiments were carried out in 11 dogs. The method exploited the relatively long tissue lifetime of Gd(ABE-DTTA), and thus no fast R1 measurement technique was needed. Myocardial perfusion was determined with colored microspheres (MP).

RESULTS

With varying extent of ischemia, impaired wall motion (WM) and lower DeltaR1(n) values were detected in the ischemic sectors, as opposed to the nonischemic sectors where normal WM and higher DeltaR1(n) were observed. Based on the DeltaR1(n), data from the myocardial perfusion assay and the DeltaR1(n) maps were compared in the ischemic sectors. A correlation analysis of these two parameters demonstrated a significant correlation (R = 0.694, P < 0.005), validating the DeltaR1(n)-mapping method for the quantitation of ischemia. Similarly, pairwise correlations were found for the MP, DeltaR1(n), and wall thickening (WT) values in the same areas. Based on the correlation between DeltaR1(n) and MP, DeltaR1(n) maps calculated with a pixel-by-pixel resolution can be converted to similarly high-resolution myocardial perfusion maps.

CONCLUSION

These results suggest that the extent of the severity of ischemia can be quantitatively represented by DeltaR1(n) maps obtained in the presence of our CA.

MR perfusion and delayed enhancement imaging in the heart

Cardiac magnetic resonance imaging is rapidly emerging as an exciting and important technique for the investigation of congenital and acquired heart disease. This article focuses in particular on recent developments in the field of adenosine stress myocardial perfusion as well as addressing the many applications of 'delayed enhancement' imaging.

Importance of Parametric Mapping and Deconvolution in Analyzing Magnetic Resonance Myocardial Perfusion Images

AIM

: We sought to improve the clinical interpretation of first-pass myocardial magnetic resonance perfusion. Parametric analyses of the myocardial distribution of the contrast agent have been proposed. The objective of the present study was to compare the effectiveness of visual analysis and of a parametric approach in an animal model under acquisition conditions as close as possible to clinical reality.

METHOD

: Experiments were conducted in vivo with various kinds of pharmacological stimulation in normal pigs and in pigs with stenosis of the left circumflex coronary artery. First-pass MR images and parametric maps were first assessed by medical experts. MR parameters, the myocardial signal intensity variation DeltaSI, ascending up-slope, and rMBF (blood flow calculated by fast discrete ARMA deconvolution) were then compared with blood flow measurements using radioactive microspheres.

RESULTS AND CONCLUSIONS

: Interobserver agreement was 57% and 81% and accuracy 53% and 81%, for visual and for parametric map analysis, respectively. For deconvolution parameters, a linear relationship y = 371 + 1.27x, r = 0.78 was obtained between rMBF calculated by ARMA and the radioactive microsphere blood flow. Moreover, the fast and robust parametric mapping of rMBF by the discrete ARMA method allows MR evaluation of myocardial perfusion independently of hemodynamic conditions.

Magnetic resonance brain perfusion imaging with voxel-specific arterial input functions

PURPOSE

To propose an automatic method for estimating voxel-specific arterial input functions (AIFs) in dynamic contrast brain perfusion imaging.

MATERIALS AND METHODS

Voxel-specific AIFs were estimated blindly using the theory of homomorphic transformations and complex cepstrum analysis. Wiener filtering was used in the subsequent deconvolution. The method was verified using simulated data and evaluated in 10 healthy adults.

RESULTS

Computer simulations accurately estimated differently shaped, normalized AIFs. Simple Wiener filtering resulted in underestimation of flow values. Preliminary in vivo results showed comparable cerebral flow value ratios between gray matter (GM) and white matter (WM) when using blindly estimated voxel-specific AIFs or a single manually selected AIF. Significant differences (P < or = 0.0125) in mean transit time (MTT) and time-to-peak (TTP) in GM compared to WM was seen with the new method.

CONCLUSION

Initial results suggest that the proposed method can replace the tedious and difficult task of manually selecting an AIF, while simultaneously providing better differentiation between time-dependent hemodynamic parameters.

Surgical management of right ventricular dysfunction late after repair of tetralogy of fallot: right ventricular remodeling surgery

With the increasing number of late survivors of repair of tetralogy of Fallot, surgical management of patients with right ventricular (RV) dysfunction and limited exercise capacity has become a more frequent problem. The wide variability in clinical status, extent of RV dilatation, and dysfunction at the time of presentation for surgical intervention has resulted in disparate surgical results after pulmonary valve insertion. With increasing use of magnetic resonance imaging, quantitative measures of RV volumes, function, and pulmonary regurgitant fraction have enabled a more systematic analysis of results. While there is a group of patients that responds favorably to pulmonary valve insertion, there is also a large subgroup that does not; this requires further analysis of the mechanisms responsible. We have developed a surgical approach to this latter group of patients, which incorporates the concepts of ventricular remodeling or restoration developed for the left ventricle following myocardial infarction. Preliminary results indicate that this procedure is equally safe to pulmonary valve insertion alone, and may result in improved RV function.

Screening with MRI: a new "all inclusive" protocol

In the past decades, only mammography has been considered to be a valuable radiological screening test in the European community, as the amount of ionizing radiation associated with other radiological imaging modalities did not justify their use for screening. MRI overcomes many limitations inherent to other imaging methods: lack of ionizing radiation, high spatial and temporal resolution, and unsurpassed soft-tissue contrast. Up to now, only few diseases have been screened for with MR: primarily breast cancer and colonic cancer. A new 60-minute comprehensive MR screening protocol is now capable of assessing the central nervous, the peripheral and the cardiovascular system, as well as the colon, allowing for screening of several diseases simultaneously.

Single-vessel coronary artery stenosis: myocardial perfusion imaging with Gadomer-17 first-pass MR imaging in a swine model of comparison with gadopentetate dimeglumine

PURPOSE

To evaluate the ability of Gadomer-17 to depict perfusion defects in a closed-chest swine model of single-vessel coronary artery disease.

MATERIALS AND METHODS

Twelve pigs underwent closed-chest placement of a flow reducer for 70%-90% luminal stenosis in the proximal left anterior coronary artery. Magnetic resonance (MR) perfusion imaging with Gadomer-17 and gadopentetate dimeglumine, microsphere blood flow (MBF) testing, and technetium 99m ((99m)Tc) 2 methoxyisobutylisonitrile (MIBI) single photon emission computed tomography (SPECT) were performed during dipyridamole vasodilation. Comparisons of percentage signal intensity (SI) increase (PSIC) in remote and ischemic myocardium were made with repeated measurements analysis of variance after injection of both tracers.

RESULTS

Perfusion defects and reduced PSIC in the anterior ischemic versus the inferior remote myocardium could be identified after injection of both Gadomer-17 (PSIC, 66% +/- 30 [mean +/- SD] vs 100% +/- 32, respectively; P <.001) and gadopentetate dimeglumine (PSIC, 49% +/- 31 vs 81% +/- 43, respectively; P <.005). The size of perfusion defect depicted with both tracers was highly correlated with defect size at (99m)Tc MIBI SPECT (r = 0.69, P <.05 for Gadomer-17 and r = 0.60, P =.05 for gadopentetate dimeglumine) and with areas of reduced MBF (r = 0.70, P <.05 for Gadomer-17 and r = 0.80, P <.05 for gadopentetate dimeglumine). PSIC also correlated with MBF (r = 0.89, P <.001 for Gadomer-17 and r = 0.75, P <.001 for gadopentetate dimeglumine). Gadomer-17 allowed differentiation of ischemic from nonischemic myocardium, as demonstrated by reduced PSIC (PSIC, 48% +/- 38 vs 72% +/- 31, respectively; P <.001) until 20 minutes after contrast material injection. In contrast, differentiation of ischemic from nonischemic myocardium was possible only until 55 seconds after injection of gadopentetate dimeglumine (PSIC, 36% +/- 24 vs 56% +/- 27, respectively; P <.005) but not at any time point thereafter.

CONCLUSION

With the study conditions, Gadomer-17 provided more prolonged differentiation of ischemic from remote myocardium than that with gadopentetate dimeglumine.

Mapping myocardial perfusion with an intravascular MR contrast agent: robustness of deconvolution methods at various blood flows

Evaluation of quantitative parameters such as regional myocardial blood flow (rMBF), blood volume (rMBV), and mean transit time (rMTT) by MRI is gaining acceptance for clinical applications, but still lacks robust postprocessing methods for map generation. Moreover, robustness should be preserved over the full range of myocardial flows and volumes. Using experimental data from an isolated pig heart preparation, synthetic MR kinetics were generated and four deconvolution approaches were evaluated. These methods were then applied to the first-pass T(1) images of the isolated pig heart using an intravascular contrast agent and rMBF, rMBV and rMTT maps were generated. In both synthetic and experimental data, the fit between calculated and original data reached equally good results with the four techniques. rMBV was the only parameter estimated correctly in numerical experiments. Moreover, using the algebraic method ARMA, abnormal regions were well delineated on rMBV maps. At high flows, rMBF was underestimated at the experimental noise level. Finally, rMTT maps appeared noisy and highly unreliable, especially at high flows. In conclusion, over the myocardial flow range, i.e., 0-400 ml/min/100g, rMBF identification was biased in presence of noise, whereas rMBV was correctly identified. Thus, rMBV mapping could be a fast and robust way to detect abnormal myocardial regions.

Assessment of hepatic perfusion parameters with dynamic MRI

Quantification of hepatic perfusion parameters greatly contributes to the assessment of liver function. The purpose of this study was to describe and validate the use of dynamic MRI for the noninvasive assessment of hepatic perfusion parameters. The signal from a fast T(1)-weighted spoiled gradient-echo sequence preceded by a nonslice-selective 90 degrees pulse and a spoiler gradient was calibrated in vitro with tubes filled with various gadolinium concentrations. Dynamic images of the liver were obtained after intravenous bolus administration of 0.05 mmol/kg of Gd-DOTA in rabbits with normal liver function. Hepatic, aortic, and portal venous signal intensities were converted to Gd-DOTA concentrations according to the in vitro calibration curve and fitted with a dual-input one-compartmental model. With MRI, hepatic blood flow was 100 +/- 35 mL min(-1) 100 mL(-1), the arterial fraction 24 +/- 11%, the distribution volume 13.0 +/- 3.7%, and the mean transit time 8.9 +/- 4.1 sec. A linear relationship was observed between perfusion values obtained with MRI and with radiolabeled microspheres (r = 0.93 for hepatic blood flow [P < 0.001], r = 0.79 for arterial blood flow [P = 0.01], and r = 0.91 for portal blood flow [P < 0.001]). Our results indicate that hepatic perfusion parameters can be assessed with dynamic MRI and compartmental modeling.

Myocardial flow reserve parametric map, assessed by first-pass MRI compartmental analysis at the chronic stage of infarction

Regional myocardial flow and flow reserve (MFR) were assessed by compartmental analysis of Gd-enhanced MRI first-pass data in 7 patients with atypical chest pain, and in 15 patients with previous transmural myocardial infarction. The FE product (Flow x Extraction coefficient), derived from the modified Kety equation, was measured in regions corresponding to the Tetrofosmine-SPECT fixed defect and in remote normal regions. The FE product at rest and hyperemic FE product were similar in healed revascularized tissues (70.5 +/- 15.6 and 112.5 +/- 19.5 ml/mn/100 g, respectively) and in normal myocardium (76.2 +/- 18.3 and 142.2 +/- 33.0, respectively). In contrast, the FE index (48.8 +/- 15.2 and 60.7 +/- 18.0, respectively, P < 0.01 versus the two previous groups) and the MFR (1.27 +/- 0.20 vs. 1.91 +/- 0.29 in normal regions) were reduced in healed fibrotic tissues when the infarct-related artery remained occluded. Myocardial flow reserve maps allowed correct identification of regions corresponding to an occluded infarct-related artery.