TussisWatch: A Smart-Phone System to Identify Cough Episodes as Early Symptoms of Chronic Obstructive Pulmonary Disease and Congestive Heart Failure

摘要: Chronic obstructive pulmonary disease (COPD) and congestive heart failure (CHF) are leading chronic health concerns among the aging population today. They both typically characterized by episodes of cough that share similarities. In this paper, we design TussisWatch , a smart-phone-based system to record process for early identification COPD or CHF. our technique, each episode, do following: 1) filter noise; 2) use domain expertise partition episode into multiple segments, indicative otherwise; 3) identify limited number audio features segment; 4) remove inherent biases as result sample size differences; 5) two-level classification scheme, based on idea Random Forests, recorded segment. Our classifier, at first-level, identifies whether not given segment indicates disease. If yes, second-level classifier symptomatic Testing with cohort 9 COPD, CHF, 18 CONTROLS subjects spread across genders, races, ages, achieves good performance in terms Sensitivity, Specificity, Accuracy, Area under ROC curve. The proposed has potential aid access healthcare, may be also used educate patients self-care home.