Three-dimension time resolved reconstruction of esophageal peristaltic contractions using simultaneous ultrasound (US) and manometry

Abstract

Background: Endohiminal ultrasound (US) images of the esophagus can be reconstructed in 3D space by moving the US transducer at a constant velocity along the esophagus, and then stacking the images using computer reconstruction algorithms. This technique was modified to evaluate peristaltic contractions in the esophagus by using a stationary probe and moving the esophagus around the probe (peristaltic contraction). Time was used as a variable (x-axis). Purpose: To evaluate peristaltic contractions in the normal esophagus using 3D time resolved reconstruction of endoluminal US images and simultaneous manometry. Method: 2D US images and simultaneous intraluminal pressure of peristaltic contractions were acquired by keeping a 20MHz high-frequency endoluminal US transducer/manometry probe at a fixed position in the esophagus of 6 normal vohimeers during swallowing. 3D time resolved US images were reconstructed using a computer-based 3D algorithm with time as the x-axis. Result: The peristaltic contraction sequence can be viewed as a 3D US image with the X-axis representing time and Y and Z axes representing space on the ultrasound images (see figure). The geometric configuration of the esophagus, the muscle thickness of the circular smooth muscle, longitudinal smooth muscle, total muscle and corresponding pressure can be evaluated during the peristaltic contraction. Conclusion: 3D time resolved reconstruction of endoluminal US images and simultaneous pressure data allow the visualization of geometric changes and correlation of muscle thickness with pressure changes during the peristalttc contraction sequence. Funded through an NIH grant R01 DK59500-01. units, a.u.), mammal relative enhancement, time of arrival of contrast inflow, time-to-peak (T-peak), wash-in (W-in), wash-out (W-out), and brevity-of-enhancement (time between points of W-in and W-out rate) (BrevEnh). Results: 5emiqnantitative data were obtained for all subjects. Comparing perfusion parameters for the pancreatic head between V vs. CPpatients showed highly significant differences for T-peak (15,4 _+ 6,1 vs 46,6 16,6 sec), W-in (99,6 39,5 vs 58,8 -+ 13,3 a.u./sec), and for BrevEnh (21,5 _+ 12,3 vs 53,3 -15,8 sec), whereas no differences were found for the other parameters measured. Discussion: MRI perfusion parameters most representative for CP were T-peak, W-in, and BrevEnFi. This is compatible with the hypothesis of an increased resistance to the normal parenchymal perfusion due to tissue ischemia. A limitation is that all patients had moderate to severe CP. The significant differences of perfusion parameters in CP vs; normal volunteers is not yet proven in mild forms of CP. Additional studies in patients with upper abdominal pain suggestive of CP, are being performed.