Initial Experience - Clinical Application and Scoring System of Real-Time Fetal Cardiac Magnetic Resonance at 0.55T (RF_FR_316)

Presenting Author(s)

• JD

  Jon A. Detterich, MD

  Professor
  Children's Hospital Los Angeles

Primary Author(s)

• JD

  Jon A. Detterich, MD

  Professor
  Children's Hospital Los Angeles

Co-Author(s)

• JH

  Jamie K. Harrington, MD

  Assistant Professor of Clinical Pediatrics
  Children's Hospital Los Angeles
• KR

  Katherine Rand, MD

  Postdoctoral Fellow
  Children's Hospital Los Angeles
• JP

  Jay D. Pruetz, MD

  Associate Director
  Guerin Family Congenital Heart Program
• 

  Ye Tian, PhD

  Research Assistant Professor
  University of Southern California
• 

  Krishna S. Nayak, PhD

  Dean's Professor
  University of Southern California
• 

  John C. Wood, MD, PhD

  Professor
  University of Southern California

Background: Fetal echocardiography is the current standard for diagnosing congenital heart disease in utero. However, poor imaging windows may hinder diagnosis when performed late in gestation, with maternal obesity, or poor fetal position. Fetal cardiac magnetic resonance imaging (FCMR) complements echocardiography, offering better image quality, especially late in gestation due to reduced fetal motion. The main barrier to clinical FCMR has been fetal cardiac gating. New FCMR techniques, such as metric-optimized or external cardiac gating systems, now enable gating to the fetal heart rate. This study describes our early experience with real-time FCMR at 0.55T.

Methods: We enrolled 29 participants to undergo FCMR with two sets of twin gestation to total 31 cardiac studies. Participants were scanned on a whole-body system (prototype 0.55T MAGNETOM Aera, Siemens Healthineers) equipped with high-performance shielded gradients (45mT/m amplitude, 200T/m/s slew rate). Imaging was performed with a real-time spiral balanced steady-state (bSSFP) pulse sequence. For each FCMR, multiple views (short-axis, 4-chamber, and 2-chamber) of the heart were captured by stacks of 2D slices. Images were reconstructed at 400 ms/frame, averaging cardiac motion while reducing intra-frame blurring due to maternal and fetal motions. We developed a grading system based on American Society of Echocardiography fetal echocardiogram guidelines and prior FCMR studies^1-5. Categorical grading of individual anatomic features was made using a 4 point scale: 1 non-diagnostic, 2 poor with artifacts, 3 good quality with artifacts, and 4 good quality without artifacts (Fig.1). Quantifiable metrics included cardiac size and axis (Fig.2).

Results: Of 31 FCMR studies, 26 agreed with echo findings (84%). Of the 5 FCMR that did not agree, 3 were abnormal by fetal echo with inconclusive FCMR findings and 2 were normal echo studies with abnormal FCMR findings (Table 1). The most common artifact was fetal motion followed by banding artifact. Mean cardiothoracic area ratio (CTA) was 27% (+/-3.8%) and cardiac axis (CAX) 45.9% (+/-7.9%). The CTA was not different between normal and abnormal FCMR exams but the CAX tended to be higher in the abnormal FCMR exams (P=0.055). Most anatomic features received a grade of 3 (good quality with artifacts). Anatomic structures receiving a lower grade (1 and 2) include atrioventricular valves, ventricular septum, semilunar valves, and pulmonary arteries. The features with the highest grade (3 and 4) were venous structures, including pulmonary veins. 

Conclusion: This is one of the first studies to qualitatively grade anatomic findings using low field fetal cardiac MRI. Our initial results demonstrate fair agreement between echo and FCMR exams with most fetal cardiac structures visualized with good quality. Certain anatomic features had poor image quality, in particular valves and pulmonary arteries, while venous structures had the highest image quality and discernability.

Figure 1 – Graphic representation of FCMR grading scale - This figure contains graphic examples of the scoring system used to evaluate cardiac structures. These examples were used specifically to assess atria and ventricles. Other anatomy seen in the grade 4 image include pulmonary left pulmonary veins, descending aorta and the IVC. The bright circle to the right of the DAO is the esophagus.

Figure 2 – Cardiac Biometry - This figure contains graphic examples of cardiothoracic area ratio and cardiac axis measurements.