Quantitative microultrasound characterization of gastrointestinal tissue for ultrasound capsule endoscopy

Abstract

Gastrointestinal (GI) disease progression is often characterized by cellular and architectural changes within the mucosal and sub-mucosal layers. One relevant disorder, Barrett’s esophagus, is of particular interest as a recognized predictor of esophageal cancer. To enhance the clinical ability to detect cellular changes deeper in tissue and earlier, we are exploring quantitative microultrasound) techniques in healthy ex vivo porcine GI tissue for implementation in ultrasound capsule endoscopy. A single-element, piezocomposite mUS transducer operating at fc = 47.7 MHz was used to obtain pulse-echo images of ex vivo porcine gastroesophageal samples, which were bisected and mechanically scanned along the long GI axis. Selected samples were mechanically separated to isolate the upper mucosal layers from the underlying muscle layers and placed on a known agar substrate to calculate the attenuation coefficient of the tissue. All other sample data were digitally segmented. Reflectivity data from the top 50, 100, and 250 μm of tissue were analysed to assess the effect of variable surface density on the calculated acoustic impedance. The entire thickness of the segmented tissue was then used to calculate position-dependent backscatter coefficient (BSC) along with intra- and inter-sample variability for use as a baseline from which to make further quantitative advances.