Opinions on Computer Audition for Bowel Sounds Analysis in Intestinal Obstruction: Opportunities and Challenges From a Clinical Point of View

Frontiers in Medicine

article

Zhu Yang¹  Luming Huang¹  Jingsun Jiang¹  Bing Hu¹  Chengwei Tang¹  Jing Li¹ 

[1] Department of Gastroenterology, West China Hospital, Sichuan University

关键词: intestinal obstruction;    bowel sounds;    computer audition;    machine learning;    deep learning;   

DOI  :  10.3389/fmed.2021.655298

学科分类：社会科学、人文和艺术（综合）

来源: Frontiers

【 摘 要 】

Intestinal obstruction (IO) is a common acute abdominal disease with abdominal pain, distension, vomiting, and constipation. IO, especially mechanical IO (MIO), may need emergency surgery, since it possibly has high morbidity and mortality in cases of perforation, intestinal fistula, peritonitis, etc. Until now, the diagnosis of IO has still relied on abdominal imaging examination (computerized tomography, X-ray), but its application is limited by its radiation, high cost, and requirement for expensive, large equipment as well as professional technicians. Bowel sound (BS) auscultation is not only safe and effective but also non-invasive for the diagnosis of IO. However, BS has strong subjectivity and randomness, and susceptibility to noises. Doctors have poor accuracy for BS auscultation, which was only 84.5% in normal people and even lower in patients with IO (only 70–80%) (1).

【 授权许可】

CC BY