Development of a bowel sound detector adapted to demonstrate the effect of food intake

Abstract

Objective Bowel sounds (BS) carry useful information about gastrointestinal condition and feeding status. Interest in computerized bowel sound-based analysis has grown recently and techniques have evolved rapidly. An important first step for these analyses is to extract BS segments, whilst neglecting silent periods. The purpose of this study was to develop a convolutional neural network-based BS detector able to detect all types of BS with accurate time stamps, and to investigate the effect of food consumption on some acoustic features of BS with the proposed detector. Results Audio recordings from 40 volunteers were collected and a BS dataset consisting of 6700 manually labelled segments was generated for training and testing the proposed BS detector. The detector attained 91.06% and 90.78% accuracy for the validation dataset and across-subject test dataset, respectively, with a well-balanced sensitivity and specificity. The detection rates evaluated on different BS types were also satisfactory. Four acoustic features were evaluated to investigate the food effect. The total duration and spectral bandwidth of BS showed significant differences before and after food consumption, while no significant difference was observed in mean-crossing rate values. Conclusion We demonstrated that the proposed BS detector is effective in detecting all types of BS, and providing an accurate time stamp for each BS. The characteristics of BS types and the effect on detection accuracy is discussed. The proposed detector could have clinical application for post-operative ileus prognosis, and monitoring of food intake.

Automated Bowel Sound and Motility Analysis with CNN Using a Smartphone

“…ML models are effective for identifying BSs [ 7 , 12 , 14 , 15 , 16 , 18 ], among which CNNs have been reported to offer better BS recognition accuracy [ 8 ]. Another neural network model that has a good reputation for sound recognition is long short-term memory (LSTM), which can process not only single data points, but the entire data sequence [ 19 , 20 ].…”

“…In addition, differences in the sound-to-sound interval (SSI) have been reported between healthy subjects and patients with IBS or Crohn’s disease, indicating that BSs can be used to diagnose bowel-related diseases and bowel motility [ 4 , 5 ]. With regard to food, BSs are also expected to be effectively used for food-intake related sensing, such as capturing changes in BSs with coffee and carbonated water intake [ 6 ] and changes in BS indices with the amount of energy contained in food [ 7 ].…”

“…BSs can be analyzed directly by a physician or using a learning model that recognizes sounds collected using a special recording instrument [ 4 , 6 , 8 , 9 , 10 , 11 ]. For instance, Craine et al [ 4 ] used a method involving stationary-nonstationary filters with wavelet transformation, whereas Wang et al [ 7 ] used an artificial neural network and Sakata and Suzuki [ 11 ] used a classification algorithm capturing sound features. Although BS measurement is not invasive and various analysis methods have been established, it requires specialized devices such as stethoscopes and contact-type recording devices that are only available in a hospital environment.…”

Practicing Digital Gastroenterology through Phonoenterography Leveraging Artificial Intelligence: Future Perspectives Using Microwave Systems

“…Bowel sound auscultation was digitally revolutionized with the development of computerized bowel sound detectors coupled to microphone-based sensors. These devices were able to adequately detect subtypes of bowel sounds with accurate start and end points [ 98 ]. Single burst (SB), multiple bursts (MB), continuous random sound (CRS) and harmonic sound (HS) patterns were recognized that were previously not detected by any device [ 98 ].…”

“…These devices were able to adequately detect subtypes of bowel sounds with accurate start and end points [ 98 ]. Single burst (SB), multiple bursts (MB), continuous random sound (CRS) and harmonic sound (HS) patterns were recognized that were previously not detected by any device [ 98 ].…”