Ralf W. Bauer MD, Presenter: Research Consultant, Siemens AG Speakers Bureau, Siemens AG

Isabel Radtke BSC, Abstract Co-Author: Nothing to Disclose

Christina Larson MD, Abstract Co-Author: Nothing to Disclose

Shuo Zhang MS, Abstract Co-Author: Research collaboration, Siemens AG

Kai Tobias Block, Abstract Co-Author: Employee, Siemens AG

Thomas G. Graf PhD, Abstract Co-Author: Employee, Siemens AG

Martin Georg Mack MD, Abstract Co-Author: Nothing to Disclose

Thomas Josef Vogl MD, PhD, Abstract Co-Author: Nothing to Disclose

The quality of cardiac MRI largely depends on the patient’s capability to hold the breath and on a good quality ECG-signal. In patients with severely impaired cardiopulmonary status both conditions may be difficult to achieve. Hence, independency from respiration and ECG-triggering is desirable. We investigated a novel TrueFISP sequence with radial k-space sampling (rTrueFISP) compared to standard balanced steady state free precision (bSSFP) TrueFISP for true real-time functional cardiac evaluation.

12 healthy volunteers (mean age 32.7 y, mean heart rate 78±12 bpm) underwent 1.5 T cardiac MRI. Single-shot short axis views (8 mm slabs, 20% gap) were acquired with a) retrospectively ECG-gated segmented bSSFP TrueFISP sequences in expiratory breath hold and with rTrueFISP sequences with sliding-window acquisition (5 subframes) with b) a temporal resolution of 15 frames/s and c) 40 frames/s. rTrueFISP were acquired during free breathing without ECG-synchronization. Left ventricular functional parameters [enddiastolic (EDV) and endsystolic volume (ESV), ejection fraction (EF)] were determined according to the Simpson’s rule with bSSFP representing the reference standard. Contrast-to-noise-ratio (CNR) of myocardium/blood pool in the left ventricle was calculated. Quality of wall motion depiction was assessed by two radiologists (1 = good, 2 = fair, 3 = non-diagnostic).

EF showed strong correlation (r = 0.91; p < 0.001) between bSSFP (65±5%) and 40 fr/s rTrueFISP (62±6%) but only moderate correlation (r = 0.67) with 15 fr/s rTrueFISP (52±6%). While EDV was in good agreement for all three sequences, ESV was significantly (p = 0.003) overestimated with 15 fr/s rTrueFISP (15 fr/s: 43.9 ml; 40 fr/s: 36.2 ml; bSSFP: 33.9 ml). Mean CNR was significantly lower (p = 0.002) with 40 fr/s rTrueFISP compared to 15 fr/s rTrueFISP and bSSFP (7.5 vs. 12.6 vs. 14.2). However, due to the faster temporal resolution image quality for wall motion assessment was rated significantly better (p = 0.03) for 40 fr/s (1.3±1.4) compared to 15 fr/s rTrueFISP (1.8±1.9).

Left ventricular functional analysis with real-time rTrueFISP is in good agreement with standard ECG-gated breath hold cine bSSFP sequences.

The independency from ECG-synchronization and breathing promises a robust method for patients with impaired cardiopulmonary status in whom breath hold and a good quality ECG-signal cannot be achieved.

Bauer, R, Radtke, I, Larson, C, Zhang, S, Block, K, Graf, T, Mack, M, Vogl, T, True Real-time Cardiac MRI in Free Breathing without ECG-Synchronization Using a TrueFISP Sequence with Radial k-Space Sampling.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11011109.html