Research Presentation Session

RPS 303 - Advanced CT and MR techniques

Lectures

1
RPS 303 - Comparison of free breathing 3D mDIXON with 3D inversion recovery and 3D spectral presaturation with inversion recovery sequences for the assessment of late gadolinium enhancement

RPS 303 - Comparison of free breathing 3D mDIXON with 3D inversion recovery and 3D spectral presaturation with inversion recovery sequences for the assessment of late gadolinium enhancement

05:25R. Dessouky, Bron / FR

Purpose:

Three dimensional (3D) late gadolinium enhancement (LGE) MRI sequences are increasingly used in clinical practice. Knowledge of strengths/limitations of each sequence can guide sequence choice. We aimed to evaluate technical and diagnostic performance of three dimensional (3D) mDIXON versus 3D inversion recovery (3D VIAB) and 3D spectral presaturation with inversion recovery (3D SPIR) LGE sequences.

Methods and materials:

A total of 78 patients (50 males and 28 females, age 49±18 years) with 1.5T CMR including 3 different 3D LGE sequences (3D mDIXON, 3D VIAB, and 3D SPIR) were evaluated for technical and diagnostic performance by two readers. Qualitative scores and quantitative signal to noise (SNR) and contrast to noise (CNR) measurements were compared among sequences. Qualitative comparisons were made using Friedman and Wilcoxon signed rank tests. Quantitative comparisons were made using one way ANOVA. Reader agreements were tested using Kohen’s Kappa. Any p-value <0.05 was significant.

Results:

Technically, 19 out of 78 (24 %) patients were excluded due to poor (grade 4) image quality and reader agreements were excellent (= 0.91-0.96, p<0.001) for all sequences. Diagnostically, 3D mDIXON showed higher confidence in the diagnosis of pericardial enhancement (p-values =0.007, 0.034 and 0.025 for three sequences, 3D VIAB, and 3D SPIR comparisons). 3D mDIXON outperformed 3D SPIR in both visualisation of LGE (p=0.02) and quality of fat suppression (p=0.001).

Conclusion:

3D mDIXON is a diagnostic problem-solving tool, especially when making a diagnosis of epicardial enhancement and/or fat suppression is needed. Choice of 3D LGE sequence should be based on a patient’s breath-hold ability, diagnostic needs, and institutional availability considering the strengths and limitations of each sequence.

Limitations:

A possible differences in timing between gadolinium injection and image acquisition.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

2
RPS 303 - Correcting versus resolving respiratory motion in free-breathing whole-heart MRA: a comparison in patients with thoracic aortic disease

RPS 303 - Correcting versus resolving respiratory motion in free-breathing whole-heart MRA: a comparison in patients with thoracic aortic disease

05:28T. Leonard, Charleston / US

Purpose:

Whole-heart magnetic resonance angiography (MRA) requires sophisticated methods accounting for respiratory motion. Our purpose was to evaluate the image quality of compressed sensing-based respiratory motion-resolved three-dimensional (3D) whole-heart MRA compared with self-navigated motion-corrected whole-heart MRA in patients with known thoracic aorta dilation.

Methods and materials:

25 patients were prospectively enrolled in this ethically approved study. Whole-heart 1.5T MRA was acquired using a prototype 3D radial steady-state free-precession free-breathing sequence. The same data was reconstructed with a one-dimensional motion-correction algorithm (1D-MCA) and an extradimensional golden-angle radial sparse parallel reconstruction (XD-GRASP). Subjective image quality was scored and objective image quality was quantified (signal intensity ratio, SIR, and vessel sharpness). Wilcoxon, McNemar, and paired t-tests were used.

Results:

Subjective image quality was significantly higher using XD-GRASP compared to 1D-MCA (median 4.5, interquartile range 4.5–5.0 versus 4.0 [2.25–4.75]; p < 0.001), as well as signal homogeneity (3.0 [3.0–3.0] versus 2.0 [2.0–3.0]; p = 0.003), and image sharpness (3.0 [2.0–3.0] vs 2.0 [1.25–3.0]; p < 0.001). SIR with the 1D-MCA and XD-GRASP was 6.1±3.9 versus 7.4±2.5, respectively (p < 0.001), while signal homogeneity was 274.2±265.0 versus 199.8±67.2 (p = 0.129). XD-GRASP provided a higher vessel sharpness (45.3±10.7 versus 40.6±101, p = 0.025).

Conclusion:

XD-GRASP-based motion-resolved reconstruction of free-breathing 3D whole-heart MRA datasets provides improved image contrast, sharpness, and signal homogeneity, and seems to be a promising technique that overcomes some of the limitations of motion correction or respiratory navigator gating.

Limitations:

A limited sample size and computational power heavy post-processing.

Ethics committee approval

IRB approved and written informed consent was obtained.

Funding:

Siemens.

3
RPS 303 - Compressed sensing 4D flow assessment of aortic and pulmonary artery flow volumes

RPS 303 - Compressed sensing 4D flow assessment of aortic and pulmonary artery flow volumes

05:59U. Reiter, Graz / AT

Purpose:

Although cardiac magnetic resonance four-dimensional (4D) flow imaging is considered an adequate technique for the assessment and “conservation of mass” comparisons of flow volumes, its widespread use is limited by long scan times. The purpose of the study was to investigate if the acceleration of whole-heart 4D flow imaging by compressed sensing (CS) can be employed to derive aortic and pulmonary flow volumes.

Methods and materials:

11 prospectively recruited cardiac patients without known or suspected shunts and with a regular heart rhythm underwent 3T (Magnetom Skyra, Siemens Healthcare, Erlangen, Germany) retrospectively ECG- and navigator-gated whole-heart 4D flow imaging with (acceleration-factor=7.6) and without CS (parallel acquisition acceleration-factor=3) employing a prototype sequence. Spatial resolution and the number of reconstructed frames were matched in the respective 4D flow protocols. For both data sets, aortic (Qa) and pulmonary (Qp) net flow volumes were evaluated using prototype software (4DFlow, Siemens Healthcare, Erlangen, Germany). The relationships between results derived with and without compressed sensing were analysed by correlation analysis; means and variances were compared employing t- and Levene test.

Results:

Scan times were significantly shorter when employing CS (369±92 s vs. 618±119 s, p<0.01). Flow volumes from CS measurements (Qa=81±20 ml; Qp=82±18 ml) had no significant bias compared to measurements without CS (Qa=83±21 ml, p=0.42; Qp=82±19 ml, p=0.80) and demonstrated strong correlations (r=0.97 and 0.95, for Qa and Qp, respectively). Pulmonary-to-systemic blood flow ratios (Qp/Qa=1.02±0.11 and Qp/Qa=1.01±0.09, with and without CS, respectively) did not differ significantly, neither in means (p=0.80) nor in variances (p=0.67).

Conclusion:

Compressed sensing allows for substantial acceleration of 4D flow acquisition without significant impact on the precision of derived aortic and pulmonary flow volumes.

Limitations:

A small patient number.

Ethics committee approval

Medical University Graz (24-126ex11/12).

Funding:

OeNB Anniversary Fund (17934).

4
RPS 303 - Value of T2 mapping in patients with chronic thromboembolic pulmonary hypertension (CTEPH) before and after balloon pulmonary angioplasty (BPA)

RPS 303 - Value of T2 mapping in patients with chronic thromboembolic pulmonary hypertension (CTEPH) before and after balloon pulmonary angioplasty (BPA)

05:40A. Hasse, Giessen / DE

Purpose:

Native T1 mapping as a marker of myocardial fibrosis showed promising correlations to right heart function and pulmonary haemodynamics in pulmonary hypertension. However, myocardial texture changes, like oedema, occur long before fibrosis arises. Therefore, the aim of this study was to assess the value of T2 mapping in patients with inoperable CTEPH before and after BPA.

Methods and materials:

MRI at 1.5 Tesla including T2 mapping and right heart catheterisation were performed before and 6 months after BPA in 30 consecutive patients (mean age 63.4 ± 10.6 years; 17 female). T2 values were measured in the right ventricular insertion points (upper and lower RVIP) and in the interventricular septum at the basal short-axis section. The results were correlated to right ventricular function (RVEF) and pulmonary haemodynamics (mPAP and PVR).

Results:

RVEF, mPAP, and PVR significantly improved after BPA (all p<0.0001). T2 values were elevated and decreased significantly after BPA (upper RVIP 61.6ms ± 6.1 to 57.0ms ± 4.5 and lower RVIP 64.4ms ± 5.6 to 58.9ms ± 4.5, p<0.0001, septum 60.2ms ± 4.0 to 58.9ms ± 3.9, p<0.02). The T2 values of the lower RVIP revealed the best correlations to RVEF, mPAP, and PVR before (r=-0.33, r=0.45 and r = 0.45 both p=0.013) and after BPA (r=-0.16, r=0.35 and r=0.14).

Conclusion:

In our cohort of CTEPH patients, septal and RVIP myocardial T2 times were elevated before BPA and decreased significantly after BPA. It is therefore highly suspected that myocardial oedema is present in mechanically demanded myocardium in pulmonary hypertension and that BPA leads to a significant oedema reduction. Therefore, we conclude that T2 mapping might enable new insights regarding therapy-response, monitoring, and prognosis.

Limitations:

A small study cohort.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

5
RPS 303 - The impact of a new deep learning denoising algorithm on LGE images

RPS 303 - The impact of a new deep learning denoising algorithm on LGE images

06:53G. Muscogiuri, Milan / IT

Purpose:

To assess the reliability of 2D segmented late gadolinium enhancement (2D-SLGE) reconstructed with a DL denoising approach of 50% and 75% of strength.

Methods and materials:

20 patients who underwent cardiac MR were enrolled in September 2019. 2D-SLGE were retrospectively reconstructed in magnitude and phase-sensitive sequences using the DL denoising approach with respectively 50% and 75%. Overall subjective image quality in both magnitude and phase-sensitive sequences with standard and DL denoising reconstructions was evaluated using a 3-point Likert scale image. Objective image quality was evaluated calculating the signal to noise ratio (SNR) and contrast to noise ratio (CNR) using different images DL denoising reconstruction. A cut off value of p<0.05 was considered statistically significant.

Results:

The overall analysis of subjective image quality was statistically significant (p:0.004), better only in magnitude images reconstructed with 75% of DL denoising compared to standard magnitude images. The CNR between scar and myocardium in phase-sensitive was not significantly different between the standard and DL denoising approach. Conversely in magnitude, the CNR improve significantly in the 50% DL denoising (p:0.01) and 75% DL denoising (p:0.02) approaches. A statistically significant improvement of SNR compared to standard magnitude reconstruction was observed for 50% DL denoising in terms of scar (p:0.02). While myocardial magnitude SNR resulted in significant improvement if compared to reconstruction in 50% DL denoising (p:0.01) and 75% DL denoising (p:0.01).

Conclusion:

The best DL denoising approach seems to be shown in magnitude reconstruction with 75% DL denoising.

Limitations:

Diagnostic accuracy should be performed.

Ethics committee approval

Still in process.

Funding:

No funding was received for this work.

6
RPS 303 - Prognostic validity of coronary flow reserve (CFR) derived from mapping-stress MRI in the risk stratification of chronic coronary syndrome patients

RPS 303 - Prognostic validity of coronary flow reserve (CFR) derived from mapping-stress MRI in the risk stratification of chronic coronary syndrome patients

06:27P. Palumbo, L'Aquila / IT

Purpose:

The need for a prognostic assessment of CAD patients is continuously increasing. Some studies have tried to validate the accuracy of novel T1 mapping application during adenosine stress in detecting coronary flow alteration. The purpose of our study was to evaluate its prognostic validity during an intermediate-to-long term follow-up.

Methods and materials:

34 patient (68,42±8,9 y; 30 males; 4 females) submitted on a 3.0T stress CMR with a CT evidence of CAD were included. Ischaemic or infarctuated CAD patients were distinguished from non-obstructive CAD patients (negative first-pass perfusion or LGE). A major adverse cardiac event and secondary cardiac outcomes were reported. The mean follow-up was 30.26±5.28 months.

Results:

All measurements were firstly repeated with two different vendors and inter-vendor agreement were calculated. Secondly, cardiac outcomes were compared with T1 values. A high incidence of cardiac outcomes was recorded in the ischaemic/infarctuated CCS patients, with no significant increase in T1 values during adenosine infusion (p-value=0,158). Among non-obstructive patients, the 56.25% referred mainly secondary cardiac outcomes. Interestingly, a lack of T1 mapping stress reactivity was found. A significant difference was found between different T1 responders and non-responders group in a Kaplan-Meier survival analysis (p-value=0,001).

Conclusion:

Our study confirms the prevalent role of stress CMR as a predictive prognostic factor in the outcome of a heart patient. T1 mapping sequences result in more sensitive and specific identification of microvascular abnormalities, increasing the clinical utility of stress CMR in identifying high-risk patients.

Limitations:

A lack of catheter-based FFR and myocardial blood flow estimation do not allow a definitive comparison with the current gold standard to confirm our findings. Moreover, the small sample size was considered with potential selection bias.

Ethics committee approval

Our study was conducted in accordance with the declaration of Helsinki.

Funding:

No funding was received for this work.

7
RPS 303 - A prospective randomised non-inferiority study of low iodine dual-energy coronary CT-angiography

RPS 303 - A prospective randomised non-inferiority study of low iodine dual-energy coronary CT-angiography

07:38D. Rotzinger, Lausanne / CH

Purpose:

To demonstrate the non-inferiority of low iodine dose dual-layer spectral detector coronary CTA (CCTA) compared to standard injection conventional CCTA for the evaluation of coronary artery disease and to compare the contrast-to-noise ratio between monochromatic and conventional CCTA reconstructions.

Methods and materials:

108 patients were randomised to undergo either a standard iodine injection protocol A (flow rate, 5 mL/s) or a reduced flow rate protocol B (flow rate, 2.5 mL/s) CCTA on a dual-layer spectral detector CT system. Conventional images only were reconstructed for protocol A. For protocol B, 55 keV monochromatic images were reconstructed. Contrast-to-noise ratio (CNR) between lumen and fat was measured in 5 coronary segments (LM, distal LAD, distal LCx, proximal RCA, and distal RCA). Two independent radiologists rated 17 coronary segments per patient regarding image quality on a 4-point Likert scale (4=best score). An examination was considered diagnostic if no segment was scored 1.

Results:

The injected contrast agent volume was 72.2±9.8 mL and 42.5±10.5 mL for protocol A and B, respectively (p<0.001). CNR was 23.3±14.2 and 28.9±18.2 in protocol A and B, respectively (p=0.015). The rate of diagnostic CCTA was 89.3% (n=50/56) and 96.2% (n=50/52) in protocols A and B, respectively. Non-inferiority of protocol B compared to A was inferred (95% CI of the difference= -0.1651 to 0.0277), with a pre-specified non-inferiority margin of 10%.

Conclusion:

Spectral CCTA with 55 keV monochromatic reconstructions allows for a 40% reduction in iodine dose compared to conventional CT angiography of the coronary arteries. Additionally, monochromatic reconstructions can improve the CNR between the coronary lumen and surrounding fat.

Limitations:

Only one monochromatic energy level (55 keV) was evaluated.

Ethics committee approval

Ethics committee approval and written informed consent were received.

Funding:

Funding was received from Bracco.

8
RPS 303 - CT-FFR profiles in patients without coronary artery disease

RPS 303 - CT-FFR profiles in patients without coronary artery disease

08:21T. Leonard, Charleston / US

Purpose:

To evaluate the effect of measurement location and lumen area changes on CT-FFR values in patients without coronary artery disease (CAD).

Methods and materials:

Patients who underwent calcium scoring (CACS) and CCTA with CT-FFR were retrospectively included. Exclusion occurred if their CACS was >0, troponin levels were elevated, or any cardiac-abnormality was found on CCTA studies. On-site CT-FFR based on an artificial intelligence, deep-learning algorithm was computed for each coronary artery at proximal, mid, and distal segments. At each location, the lumen area and Hounsfield Unit (HU) value were measured. CT-FFR was considered positive with values <0.75. The relationship between lumen areas, HU-values, and CT-FFR was evaluated for each coronary artery and each location. Ratios between mid and distal values compared to proximal-values for lumen and HU parameters were calculated.

Results:

106 patients were included. In 39 (37%) patients, the LAD had CT-FFR values <0.75, with a decrease in CT-FFR from 0.97 (SD:0.04) proximally to 0.62 (SD:0.10) distally in the abnormal patients. The Cx showed a limited number of patients with CT-FFR values <0.75 (n=16, 15%), with a decrease in CT-FFR values from 0.96 (SD:0.04) proximally to 0.65 (SD:0.09) distally in those patients. The RCA had 36 (34%) patients with CT-FFR <0.75, with distal CT-FFR values of 0.61 (SD:0.12) and proximal CT-FFR values of 0.98 (SD:0.02). 12 abnormal CT-FFR values were measured mid-segment, while all others were measured at distal-segments. The lumen area was not significantly different between the abnormal and normal CT-FFR groups, while both HU and HU-ratios were significantly lower in the abnormal CT-FFR group for all three major coronary arteries.

Conclusion:

CT-FFR values in patients without CAD can become abnormal at a distal location without indicating flow-limiting stenosis, which depends strongly on HU-values.

Limitations:

A retrospective single-centre study with limited subject-availability.

Ethics committee approval

Institutional-Review-Board approval.

Funding:

Siemens Healthineers.

9
RPS303 - 4D-flow assessment of early-diastolic mitral annular peak tissue velocity: a comparison with echocardiography

RPS303 - 4D-flow assessment of early-diastolic mitral annular peak tissue velocity: a comparison with echocardiography

05:55C. Reiter, Graz/AT

Purpose:

Early-diastolic mitral annular peak tissue velocities represent important parameters in the evaluation of left ventricular diastolic function. Their measurement is well-established in echocardiography but not in cardiac magnetic resonance (CMR) imaging. The purpose of the present study was to investigate the feasibility of derivation of early-diastolic mitral annular peak tissue velocities from CMR four-dimensional (4D) flow imaging and to compare the results with echocardiography.

Methods and materials:

74 subjects (male/female, 33/41; age 60±10 years) were prospectively recruited to undergo echocardiography and CMR whole heart 4D-flow imaging at 3 Tesla. Echocardiographic medial (e’ medial) and lateral (e’ lateral) early-diastolic mitral annular peak tissue velocities were acquired following current guidelines. Corresponding 4D-flow-based early-diastolic mitral annular peak tissue velocities were determined as early-diastolic peak mean through-plane tissue velocities in the inferoseptal (4DFe’ medial) and the anterolateral (4DFe’ lateral) segments of multiplanar reformatted short-axis planes just below the mitral valve using a prototype software (4DFlow, Siemens Healthcare, Erlangen, Germany). A comparison of CMR and echocardiography data was performed by means of correlation and a Bland-Altman analysis.

Results:

Tissue velocities from 4D-flow measurements (4DFe’ medial=7.4±2.3cm/s; 4DFe’ lateral=9.3±2.5cm/s) had no significant bias compared to echocardiographically obtained tissue velocities (e’ medial=7.6±2.4cm/s, p=0.43; e’ lateral=9.3±2.9cm/s, p=0.95) while showing strong correlations (r=0.77 and 0.84, for medial and lateral assessment, respectively). The clinically frequently used average of medial and lateral velocities showed even stronger correlation (r=0.90) without bias (4DFe’ average=8.4±2.2cm/s vs. e’ average = 8.5±2.4cm/s, p=0.52).

Conclusion:

Early-diastolic mitral annular peak tissue velocities from echocardiography and CMR 4D-flow correlate strongly and demonstrate no bias, suggesting 4D-flow as a potential technique for its assessment interchangeable with echocardiography.

Limitations:

A single observer and single-centre study.

Ethics committee approval:

NCT01728597, NCT03253835.

Funding:

OeNB-Anniversary-Fund Nr.17934.

PEP Subscription Required

This course is only accessible for ESR Premium Education Package subscribers.