Evaluation of bile reflux in HIDA images based on fluid mechanics

• We used optical flow to ascertain the direction and velocity of bile flow to help physicians diagnose bile reflux exact.
• We combined image processing technology and hydrodynamic models to quantify bile reflux.
• We reconstructed 3D images from SPECT and CT to help physicians diagnosis.

We propose a new method to help physicians assess, using a hepatobiliary iminodiacetic acid scan image, whether or not there is bile reflux into the stomach. The degree of bile reflux is an important index for clinical diagnosis of stomach diseases. The proposed method applies image-processing technology combined with a hydrodynamic model to determine the extent of bile reflux or whether the duodenum is also folded above the stomach. This condition in 2D dynamic images suggests that bile refluxes into the stomach, when endoscopy shows no bile reflux. In this study, we used optical flow to analyze images from Tc99m-diisopropyl iminodiacetic acid cholescintigraphy (Tc99m-DISIDA) to ascertain the direction and velocity of bile passing through the pylorus. In clinical diagnoses, single photon emission computed tomography (SPECT) is the main clinical tool for evaluating functional images of hepatobiliary metabolism. Computed tomography (CT) shows anatomical images of the external contours of the stomach, liver, and biliary extent. By exploiting the functional fusion of the two kinds of medical image, physicians can obtain a more accurate diagnosis. We accordingly reconstructed 3D images from SPECT and CT to help physicians choose which cross sections to fuse with software and to help them more accurately diagnose the extent and quantity of bile reflux.