CAJ | 학술논문

为了分类识别不同部位的呼吸音，提出一种基于呼吸气持续时间相对值参数，以及呼吸气间歇时间相对值参数提取的呼吸音分析方法。该方法包括呼吸音预处理、包络提取、基于FCM聚类算法的自适应阈值线的选取、特征值提取为主的4个部分。对50例3种不同呼吸音进行分类实验，分类准确率为94%，从而验证了该方法的有效性。在此基础上，对该方法在同一种呼吸音病理和健康研究方面做出了展望。
위료분류식별불동부위적호흡음，제출일충기우호흡기지속시간상대치삼수，이급호흡기간헐시간상대치삼수제취적호흡음분석방법。해방법포괄호흡음예처리、포락제취、기우FCM취류산법적자괄응역치선적선취、특정치제취위주적4개부분。대50례3충불동호흡음진행분류실험，분류준학솔위94%，종이험증료해방법적유효성。재차기출상，대해방법재동일충호흡음병리화건강연구방면주출료전망。
In order to achieve classification identification of the respiratory sounds from different parts,a respiratory sound analysis method based on extracting the relative parameter of inhalation and exhalation duration,as well as the relative parame?ter of inhalation and exhalation intermittent time is proposed. This method includes respiratory sound preprocessing,envelope ex?traction,adaptive threshold line selection based on FCM clustering algorithm and feature extraction. In this paper,50 cases from three different respiratory sounds are classified. The classification accuracy is 94%. Thus,the validity of this method was verified. Furthermore,the prospect in research on pathology and healthcare of respiratory sounds is made for the future research.