1356675 - Comparison of Accelerated Compressed Sensing 2D Phase Contrast vs. Conventional Phase Contrast Technique for Assessment of Great Vessels Flow

Friday, January 27, 2023

8:42 AM – 8:49 AM

Location: Shorebreak

Presenting Author(s)

mE

mariana l. Elhage, RT

cardiac mri technologist
Houston Methodist DeBakey Heart & Vascular Center
houston, Texas, United States

Primary Author(s)

mE

mariana l. Elhage, RT

cardiac mri technologist
Houston Methodist DeBakey Heart & Vascular Center
houston, Texas, United States

Co-Author(s)

jessica Lowe, RT

lead cardiac mri technologist
Houston Methodist DeBakey Heart & Vascular Center
houston, Texas, United States
NJ

Ning Jin, PhD

Senior Key Expert
Siemens Medical Solutions USA, Inc., Ohio, United States
DN

Duc T. Nguyen, MD

ACDM Senior research associate
Houston Methodist DeBakey Heart & Vascular Center, United States
EG

Edward A. Graviss, PhD, MPH

Scientist
Houston Methodist DeBakey Heart & Vascular Center, Texas, United States
Andrada C. Guta, MD, PhD

Postdoctoral research fellow
Houston Methodist DeBakey Heart & Vascular Center
Houston, Texas, United States
Dipan J. Shah, MD

Chief, Division of Cardiovascular Imaging
Houston Methodist DeBakey Heart & Vascular Center
Houston, Texas, United States

Disclosure(s):

Andrada C. Guta, MD, PhD: No financial relationships to disclose

Dipan J. Shah, MD: No relevant disclosure to display

Background: Two-dimensional phase contrast cardiac magnetic resonance is a robust technique for assessment of flow, but can require long breath hold times, which can be cumbersome in frail patients. Compressed sensing (CS) phase contrast (PC) technique has been previously studied in a pre-clinical setting and has proven to be a useful tool in decreasing PC sequence time (Liang, 2009). This study is aimed to evaluate the agreement of the CS-accelerated and conventional PC for flow quantification in a clinical setting.

Methods: A total of 17 consecutive patients (11 inpatients) undergo both conventional PC and a novel CS PC research application. Aortic (AO) and pulmonic (PA) flows were obtained using both conventional PC sequences and the novel CS PC prototype, using a 1.5 Tesla clinical scanner (MAGNETOM Sola, Siemens Healthcare, Erlangen, Germany). The parameters for the conventional vs CS PC were the following: acceleration factor of 2 vs 7.2 [R] and spatial resolution of 1.7 x 1.7 mm vs 2 x 2 mm, with a 6 mm slice thickness for both. Temporal resolution was 27 ms for both techniques. Manual regions of interest were drawn around each great vessel and used to quantify flow (Figure 1).

Results:

Median (interquartile range, IQR) age in the cohort was 55 (51, 67) years, with 59% females. Median (IQR) heart rate was 78 (63, 88) bpm and 15 patients received gadolinium contrast before acquisition. The mean acquisition time was significantly reduced using the novel CS PC sequence compared to standard flow sequence: 4 sec (5 beats) vs 14 sec (19 beats) for both the AO and PA flow. Median AO and PA flow using CS PC vs standard PC sequence are detailed in Figure 2 (Bland-Altman plots). There was a good agreement between the two sequences for the AO flow (mean difference -0.17, 95% limits of agreement -1.3, 1.0, p = 0.12), with a slightly bias for the PA flow (mean difference -0.34, 95% limits of agreement -1.8, 1.1, p = 0.01).

Conclusion: The CS 2D flow sequence shows good agreement with minimal bias in determining AO and PA forward flow, requiring only 25% of the time needed for the conventional PC sequence. This technique could become a widespread tool, not only in cases where breath holds are challenging but as a strategy to reduce overall scan duration and improve patient comfort.