Cough Classification Tool for Early Detection and Recovery Monitoring of Tuberculosis and Asthma

Now-a-days there is significant increase in research for acoustic signal analysis in prospect of healthcare domain, one such area of interest is cough signal analysis which is major symptom of chronic diseases like tuberculosis and asthma. Cough can appear sporadically with common illnesses (e.g. cold), but when it becomes chronic in the case of asthma and tuberculosis which severely impair life quality. Though there are existing systems in analyzing cough signal, still there is a need in developing tool which analyses cough signal and capable of detecting the disease at earlier stages as well as monitoring the recovery of patients suffering either tuberculosis or asthma. The software tool proposed will take cough sounds as input and classifies them as wet or dry cough by features MFCC and classification algorithm K-Means with an efficiency of 82%. The classification will further be used as part of the tool being developed for monitoring the recovery of the diseases like tuberculosis, asthma. KeywordsWet Cough, Dry cough, MFCC Kmeans, chronic, tuberculosis, asthma I.INTRODUCTION Cough is one of the most common symptoms among all respiratory diseases. The main activity of the cough is to clear the breathing airways from foreign objects, secretions and mucus. Depending on how often it occurs and its severity, cough may persist and become chronic in nature. One third of the pulmonologist consultations are chronic cough diseases. Primarily coughs are of two kind’s wet and dry cough. Differentiation of wet and dry coughs are very subjective, it may be difficult for patients to describe their cough symptoms to doctor, which makes the medication and treatment difficult at times , so there is significant focus in acoustic signal analysis for healthcare in early detection and monitoring. In this paper we discuss about the tool developed. This tool takes cough signal as input and classifies as wet/dry cough, Further the tool helps in monitoring the disease recovery for tuberculosis and asthma and even the early detection of the disease mentioned above .This paper is organized as follows. II.CHARACTERISTICS OF COUGH SOUND Usually cough sounds consists of three phases: phase1: initial burst, phase 2: noisy air flow and phase 3: glottal closure A. Dry cough characteristics Dry cough does not contain any mucus or sputum and these three phases of cough are visible in the dry cough as shown in the Figure 1(a), observe that after the phase 1 i.e. initial burst, the energy levels are low in phase at higher frequencies. Figure 1. a) Dry cough signal with 3 phases: Phase 1, Phase 2, and Phase 3, b) Spectrogram of dry cough signal. B. Wet cough characteristics Wet cough is produced by the bacteria or virus causing secretion in the lower airways: bronchitis, asthma etc. therefore the wet cough consists of mucus and sputum. Figure 2(a) represents the wet cough; the wet cough consists of higher energies in the phase 2 Dr.Y.PADMA SAI , S.V.N NARAYANA RAO, V NAVEEN KUMAR, V PRIYANKA BRAHMAIAH,D AJAY Electronics and Communication Department VNR VIGNANA JYOTHI INSTITUTE OF ENGINEERING AND TECHNOLOGY Bachupally, Nizampet (S.O), Hyderabad, Telangana, India-5000090