連續肺音聽診系統開發

Abstract

呼吸道疾病是全球第三大死因，且在經歷重大手術或是插管呼吸器治療時，肺部及心因性併發症是醫療成本與死亡率上升的兩個重要原因，但要正確評估處置肺部疾病卻是長年以來在取決於人力成本與病況惡化間的難解議題。本研究以此為動機，結合臨床經驗與意見，研發一種達成密集持續監測，又可符合臨床人員工作流程的可連續性監測紀錄肺音聽診系統。除將研究成品與類似競品以聲學實驗方式實施頻響測試以驗證收音表現外，並根據手術中及加護病房之臨床聽診操作，設計良好的臨床肺音紀錄方法，可收集更適於機器學習或大數據分析之臨床肺音數據，有助於提升醫療照護品質。

Respiratory diseases are the third leading cause of death in the world. In major surgeries and endotracheal intubation ventilation therapies, pulmonary and cardiogenic complications are two primary causes of rising medical costs and mortality increasing, but the proper evaluation and treatment of lung diseases has been a long-term debated topic, and also a difficult problem between the cost of labor and the progression of disease. As a motivation to solve the problem, this study combined clinical experience and opinions to develop a lung sound auscultation system capable of continuous monitoring and recording, which achieves intensive continuous monitoring and meets the clinical staff's workflow at the same time. This research put the finished system and similar competing products in the acoustic test to verify its audio performance, and designed a proper clinical lung sound recording method from the clinical auscultation operation in operation rooms and intensive care units. The data recorded by this system proved to be more suitable for either machine learning or big data analysis in clinical lung sound data, and the result can be expected to help improve the quality of medical care.