Chronic obstructive pulmonary disease prediction using Internet of things-spiro system and fuzzy-based quantum neural network classifier

Abstract

Chronic Obstructive Pulmonary Disease (COPD) is a multifarious progressive disease that increases the mortality and morbidity ratio as well as becoming a life-threatening issue of an individual. Accurate and cost-effective diagnosis of diseases plays a primary role in the medical domain and a wide range of research has been carried out on disease prediction using sensory approaches along with the assistance of machine learning techniques. The traditional disease diagnosis procedures are invasive, costlier and the decision support systems were unreliable most of the time. The human exhaled breath discharged from the body is composed of various Volatile Organic Compounds (VOCs) which can be influenced by metabolic and disease activities. Hence, the analysis of VOCs in exhaled breath has an incredible potentiality for COPD diagnosis and can rapidly decrease the mortality rate. In this research, IoT-Spiro System is designed and an intelligent machine learning forecasting framework (IMLFF) has been proposed. IoT-Spiro System perceives the various VOCs patterns available in exhaled breath and that real-time parameter has been analyzed using IMLFF. The proposed framework incorporates a hybrid Genetic Big Bang-Big Crunch (GBB-BC) algorithm for selecting the optimal features from the real-time dataset and a Fuzzy-based Quantum Neural Network (F-QNN) classifier for diagnosing COPD. The experimental results illustrate that IMLFF outperforms when compared to recent existing approaches concerning various statistical parameters and performance metrics. From the result analysis, it has been determined that IoT-Spiro System and IMLFF framework can serve as an efficient assisting model to the medical practitioner for diagnosing COPD.