RPS 2303 - Advanced techniques in cardiac MR

Author Block: A-M. Vuorinen, L. Lehmonen, J. Karvonen, M. Holmström, S. Kivistö, T. Kaasalainen; Helsinki/FI
Purpose or Learning Objective: Cardiac implantable electronic devices (CIED) induce artefacts on MRI and may significantly reduce the diagnostic value of cardiac magnetic resonance imaging (CMR). The study aimed to assess the effect of CIED generator location and a raised-arm imaging position on the CIED-induced artefacts on CMR.
Methods or Background: All clinically indicated CMRs performed in our institution for CIED patients with normal cardiac anatomy and permanent CIED with endocardial pacing leads between November 2011 and October 2019 were included (n=171). Cine and late gadolinium enhancement (LGE) images were analysed according to the American Heart Association 17-segment model for artefacts.
Results or Findings: Right-sided generator implantation and raised-arm imaging were associated with a significantly increased number of artefact-free segments. In patients with a right-sided pacemaker, the median percentage of artefact-free segments in short-axis balanced steady-state free precession LGE was 93.8% (IQR 9.4%, n=53) compared to 78.1% (IQR 20.3%, n=58) with a left-sided pacemaker (p<0.001). In patients with a left-sided implantable cardioverter defibrillator, the median percentage of artefact-free segments was 87.5% (IQR 6.3%, n=9) with raised-arm imaging, compared to 62.5% (IQR 34.4%, n=9) with arm down in spoiled gradient-echo short-axis cine (p=0.02).
Conclusion: Raised-arm imaging is a simple, costless method for reducing CMR artefacts in patients with left-sided CIED, and can be used with other image quality improvement methods. Right-sided generator implantation could be considered in patients who are known to require subsequent CMRs to ensure adequate image quality.
Limitations: Assessing artefacts included subjective judgement. Artefacts were not evaluated by specific CIED models.
Ethics committee approval: The study was approved by the Helsinki University Hospital Medical Imaging Center review board.
Funding for this study: This study was supported by HUS Medical Imaging Center research grant and Ida Montin Foundation research grant.

Author Block: B. Longere¹, N. Abassebay¹, S. Toupin², C. V. Gkizas¹, A. Simeone¹, J. Hennicaux¹, M. Schmidt³, J. Pang⁴, F. Pontana¹; ¹Lille/FR, ²Saint-Denis/FR, ³Erlangen/DE, ⁴Chicago, IL/US
Purpose or Learning Objective: The image quality of cine imaging is highly affected by arrhythmia and shortness of breath. This study aimed to evaluate whether a new compressed sensing free-breathing cine sequence (CS-FB) with motion correction (MOCO) can address such limitations.
Methods or Background: Forty-eight patients referred for cardiac MRI underwent both conventional bSSFP multi-shot multi-breath-hold cine (BH-SSFP, GRAPPA=2) and prototype free-breathing single-shot bSSFP cine (CS-FB, CS acceleration=9.0-24.0, depending on heart rate) with fully automated MOCO. Short-axis stacks were acquired for each patient with matched parameters. Acquisition and reconstruction times, image quality (Likert scale from 1 to 4), left ventricular (LV) and right ventricular (RV) volumes, LV, and RV ejection fractions (EF) and LV mass were assessed for each sequence.
Results or Findings: CS-FB cine’s scan time was shorter (2.2±0.6 min vs 4.8±1.2 min for BH-SSFP, p<0.0001), although reconstruction times were longer for CS-FB (4.7±1.5 min vs 0 min for BH-SSFP, p<0.0001). CS-FB achieved higher image quality (3.9±0.3) than BH-SSFP (3.6±0.6, respectively, p=0.0001), especially for patients with arrhythmia or difficulties holding breath (n=19; 3.9±0.2 vs 3.3±0.7, p=0.002). Regardless of cardiac rhythm, LVEF, LV stroke volumes and mass were similar between both sequences. There was difference in RVEF (CS-FB: 49.9±11.2%; BH-SSFP: 51.5±11.4%; p=0.0007) and RV stroke volume (CS-FB: 73.9±24.4mL;BH-SSFP: 76.8±26.3mL; p=0.005) but not for the RV end-systolic and end-diastolic volumes.
Conclusion: Free-breathing CS acquisition with MOCO can be reliably performed in a clinical setting to evaluate LV parameters, with better image quality and shorter scan time than conventional cine, especially for patients with arrhythmia or shortness of breath.
Limitations: CS-FB sequence required longer reconstruction times than the reference BH-SSFP cine. However, the robustness of the CS-FB sequence against arrhytmia and respiratory motion avoid the repetition of cine acquisitions.
Ethics committee approval: The ethics committee approval was obtained, IRB number: CRM-2103-163.
Funding for this study: No funding was received for this study.

Author Block: H. Chen, A. Kisters, M. K. Świderska, E. Cavus, D. Säring, K. Müllerleile, G. K. Lund, G. Adam, E. Tahir; Hamburg/DE
Purpose or Learning Objective: Cardiac adaptation in endurance athletes is a well-known phenomenon, but the acute impact of strenuous exercise is fairly unknown. The purpose of this study was to analyse the alterations in biventricular and biatrial function in triathletes after an endurance race using novel feature-tracking cardiac magnetic resonance (FT-CMR).
Methods or Background: Fifty consecutive triathletes (45±10 years; 80% men) and twenty-eight controls were prospectively recruited. All underwent 1.5T CMR examination. The time interval between race completion and CMR was 2.3±1.1 hours (range 1-5 hours). Biventricular and biatrial volumes, left ventricular ejection fraction (LVEF), FT-CMR analysis and late gadolinium enhancement (LGE) imaging were performed. Global systolic longitudinal strain (GLS), circumferential strain (GCS) and radial strain (GRS) were assessed. CMR was performed at baseline and following an endurance race. High-sensitive Troponin T and NT-proBNP were determined.
Results or Findings: Post-race Troponin T (P<0.0001) and NT-proBNP (P<0.0001) were elevated. LVEF remained constant (62±6 vs 63±7%, P=0.607). Post-race LV GLS decreased by tendency (-18±2 vs -17±2%, P=0.054), whereas GCS (-16±4 vs -18±4%, P<0.05) and GRS increased (39±11 vs 44±11%, P<0.01). Post-race right ventricular GLS (-19±3 vs -19±3%, P=0.668) remained constant and GCS increased (-7±2 vs -8±3%, P<0.001). Post-race left atrial GLS (30±8 vs 24±6%, P<0.0001) decreased while right atrial GLS remained constant (25±6 vs 24±6%, P=0.519).
Conclusion: The different alterations of post-race biventricular and biatrial strain might constitute an intrinsic compensatory mechanism following an acute bout of endurance exercise. The combined use of strain parameters may allow a better characterisation and understanding of ventricular and atrial function in endurance athletes.
Limitations: This study had a relatively small sample size and ruled out triathletes with pre-existing cardiovascular disease and systemic diseases, limiting its generalisability.
Ethics committee approval: The ethics committee approved the study (PV4764).
Funding for this study: No funding was received for this study.

Author Block: J. Liu¹, J. Li¹, H. Pu¹, W. He¹, N. Tong¹, X. Zhou², l. PENG¹; ¹Chengdu/CN, ²Shanghai/CN
Purpose or Learning Objective: Obesity increases the risk of heart failure and cardiovascular mortality. However, the specific effects of obesity on cardiac structure and function have not been clarified. This study aimed to evaluate the left ventricular (LV) geometric and functional changes using cardiovascular magnetic resonance (CMR) in uncomplicated obese adults.
Methods or Background: A total of 48 obese participants and 25 healthy controls were prospectively enrolled. The LV geometry, global systolic functions, and strains were assessed using CMR. Body composition was measured using dual X-ray absorptiometry.
Results or Findings: Compared with healthy controls, the obese patients had greater LV size, mass, myocardial thickness, and impaired myocardial contractility with lower LV global radial, circumferential, and longitudinal peak strains (PS) and LV global circumferential and longitudinal peak diastolic strain rates (PDSR) (all p<0.05). Multivariable linear regression showed that the body mass index (BMI) was independently associated with LV average myocardial thickness (LVAMT), LV maximum myocardial thickness (LVMMT), and concentricity in the obese group (LVAMT: β=0.098, p=0.02; LVMMT: β=0.16, p=0.044; concentricity: β=0.013, p=0.037). In addition, the visceral adipose tissue (VAT) was associated with LV global longitudinal PS and LV longitudinal and circumferential PDSR (longitudinal PS: β =-2.784, p<0.001; longitudinal PDSR: β=-0.202, p=0.001; circumferential PDSR: β=-0.193, p=0.005).
Conclusion: LV geometric remodelling and subclinical dysfunction are observed in obese adults with preserved LV ejection fraction. Instead of BMI, VAT is found to be a sensitive predictor for subclinical LV dysfunction.
Limitations: A cross-sectional study.
Ethics committee approval: The study was approved by the Institutional Review Board of West China Hospital.
Funding for this study: This work was supported by the National Natural Science Foundation of China [Grant number.81601462].

Author Block: X-m. WU, X-y. WU, h. tan, L-f. MA, Q. JIANG, H-p. DONG; Shanghai/CN
Purpose or Learning Objective: The objective of this study was to investigate the possibility of cardiac DWI using conventional spin-echo (SE) EPI.
Methods or Background: Seven female and five healthy male subjects (mean age, 39.2 ± 13.0 years; heart rate, 69.8 ± 7.8 bpm) without any contraindication to MR scanning underwent cardiac DWI, which was performed on a 3 Tesla MR scanner (Elition X, Philips Healthcare, Best, the Netherlands, Max G 45 mT/m, Max SR 200 T/m/s) with a 16-channel anterior coil and a 12-channel posterior coil. Diffusion-weighted SE-EPI with respiratory navigation, volume shim, and fat suppression (both SPIR and gradient reversal) was then performed with the following parameters: voxel size = 2.5 × 2.5 × 8 mm3, FOV 320 × 320 mm2, slice number 1, half scan factor 0.6, SENSE factor 2.0, NSA 10, TR 1 heartbeat, 5 b values (200, 300, 500, 800, 1000 s/mm2) were acquired separately with b0, corresponding TEs are 35, 38, 42, 47, 50 ms. Scan time ~1min6s for each b value scan considering navigation efficiency of 60%.
Results or Findings: Image quality degraded, and distortion became more severe, with higher b values. When b value > 500 s/mm2, the overall mean image quality score was lower than 2.5.A significant difference was also found between segments (SNR: F=4.59, P=0.001; CNR: F=4.60, P=0.001). The signal was lower in the interventricular septum, especially in the infero-septal part, compared with other segments.
Conclusion: Using a clinical 3T scanner, SE-EPI DWI could be successfully performed with careful choice of navigator placement and cardiac trigger delay. Medium b values are recommended to obtain good image quality.
Limitations: Only healthy volunteers were involved in this study. Patients should be included to investigate the diagnostic value of cardiac DWI in future studies.
Ethics committee approval: Not applicable.
Funding for this study: Not applicable.

Author Block: M. C. Halfmann¹, T. S. Emrich², K-F. Kreitner¹; ¹Mainz/DE, ²Charleston, SC/US
Purpose or Learning Objective: Development of compressed sensing (CS) cardiac magnetic resonance (CMR) sequences rapidly accelerated cardiac imaging while showing good reproducibility for volumetric analyses. However, for imaging biomarkers such as strain, CS-derived measurements are significantly altered. Novel prototype rapid strain parameters (junction strain (JS) and long axis strain (RS)) are evaluated for their potential to overcome that bias.
Methods or Background: 15 prospectively enrolled healthy volunteers (HV) underwent CMR including a stack of short-axis slices and two orthogonal left ventricular long-axis slices. Acquisitions were based on balanced steady-state free precession Cine (bSSFP) and CS-sequences. Exclusion criteria were any history of cardiac disease and abnormal volumes as characterised by CMR. Dedicated post-processing software (cvi42 Circle) was used to compute JS, RS and global longitudinal strain (GLS).
Results or Findings: GLS correlated moderately between bSSFP and CS (r=0.58). JS and RS correlated strongly between acquisitions (r= 0.70/0.69, respectively). Correlation between novel parameters and GLS was moderate in bSSFP (JS/GLS r=0.56; RS/GLS r= 0.53) and weak in CS (JS/GLS r=0.34; RS/GLS r= 0.31). Bland-Altman analysis showed a significant bias for CS-derived GLS with a mean difference of -3.9% and limits of agreement (LoA) from -6.9 to -1.0%. There was no significant bias for both JS and RS (mean differences -1.1/-1.0%; LoA -3.8 to 1.7/-3.8 to 1.8%).
Conclusion: Novel rapid strain parameters correlate strongly between bSSFP and CS acquisitions and with established GLS. However, they differ from GLS as there is no significant systematic bias between acquisition types.
Limitations: Rapid post-processing strains are still under evaluation, larger studies including patient populations are needed to evaluate the clinical significance of these preliminary results.
Ethics committee approval: The study protocol was approved by the local ethics committee with a waiver for informed consent.
Funding for this study: No outside funding was acquired for this study.