Early Detection of Chronic Obstructive Pulmonary Disease (COPD) Using Logistic Regression: A Machine Learning Framework for Spirometry Data

Abstract

COPD is a long term pulmonary disorder that requires early diagnosis in order to be managed. This research paper presents the Logistic Regression-based model of COPD early detection on the basis of spirometry values, such as Forced Expiratory volume in 1 second (FEV1) and Forced Vital Capacity (FVC) along with demographic and clinical characteristics. High accuracy of the model was observed, 92.5% and AUC of 0.96, which showed that the model was highly discriminative; it can discriminate between people who are healthy and those with COPD. The possibility of integrating clinical and demographic information with spirometry data in the framework of COPD classification enables the classification of COPD, even during its early phases, with high precision. The findings outline the possibility of using this method to optimize the preventive programs of early detection, decrease the healthcare load, and ameliorate the patient quality of life by providing them with preventive measures in a timely manner.