UCL Discovery

Abstract

Ascites is more difficult to detect when only a small quantity is present. The aim of this pilot study was to determine the optimal bowel sound characteristics in order to distinguish no ascites from small-volume ascites by advanced processing of bowel sound wave patterns. This analysis results in the definition of the normal range of bowel sound patterns, thus providing a novel, simple, and noninvasive way of determining on abnormal pattern, which may reflect presence of small volume ascites. Cirrhotic patients with radiologically proven small-volume ascites and a control group were subjected to bowel sound recordings. The latter were analyzed using a denoising wavelet transform-based filter and a higher-order crossings-based technique in a blinded fashion for linearly distinguishing the two classes. Scatter plots of third-order zero crossings reflect distinct changes seen in the denoised bowel sound pattern between patients and controls due to altered transmission path, providing a distinct separation of all cirrhotic patients with small ascites from controls (P < 0.0001). We conclude that the proposed bowel sounds analysis appears to provide new information regarding the changes of the bowel sound patterns due to the presence of small-volume ascites, potentially contributing towards a safe, effective, noninvasive, and easily implemented alternative method for the diagnosis of small volume ascites at the bedside.