Dynamic MR imaging of the gastroesophageal junction in healthy volunteers during bolus passage.

Purpose: To evaluate the feasibility of noninvasive dynamic fast magnetic resonance imaging (MRI) during swallowing in healthy volunteers, and to determine esophageal function at the gastroesophageal junction during swallowing. Materials and Methods: A total of 20 healthy volunteers underwent MRI while swallowing in the supine position. Dynamic gradient-echo (GRE) sequences (balanced fast field echo [B-FFE]) were employed in three planes on a 1.5T unit using a phased-array body coil. Buttermilk spiked with gadolinium (Gd) chelate (40:1) for bolus passage was used as an oral contrast agent. We evaluated visualization of esophageal bolus transit, bolus transit time (BTT), peristalsis, identification of the gastroesophageal junction, and reflux during the Valsalva maneuver. Results: The mean visible length of the esophagus was 16.2 +/- 5.3 cm, in the sagittal view, and 13.8 +/- 4.9 cm in the coronal view. In the sagittal view the BTT was defined in 15 of 20 volunteers and was 7.6 +/- 1.4 seconds. The BTT in the coronal view was measured in seven of 20 volunteers and was 8 +/- 1.3 seconds on average. The axial view yielded higher scores (2.25) than the coronal (1.98) and sagittal (1.78) views for identification of the cardia and during the Valsalva maneuver. Bolus contrast was better displayed in the sagittal (2.2) view than in the coronal (2.08) or axial (1.73) planes. In six volunteers, gastroesophageal abnormalities, such as axial hernia, reflux, and nonperistaltic contractions, were identified. For statistical analysis we used the Friedman test and a one-way analysis of variance (ANOVA). Conclusion: The results indicate that dynamic MR swallowing is a feasible and reproducible technique that warrants further studies in patients.