Detection of Early Lung Cancer Cases in Patients with COPD Using eNose Technology: A Promising Non-Invasive Approach

The integration of electronic nose (eNose) technology and advanced machine learning methodologies presents an innovative pathway for the non-invasive differentiation of chronic obstructive pulmonary disease (COPD) and lung cancer. This study utilizes a group of 84 COPD patients and 53 lung cancer patients to evaluate the effectiveness of a developed classification model. Through the cross-validation, the model displays outstanding accuracy, achieving 94.16% precision in distinguishing between these two respiratory conditions. Impressively, the model demonstrates sensitivity and specificity rates of 96.34% and 90.91%, respectively. The model's robust performance is measured by an impressive Area Under the Curve (AUC) value of 0.91, highlighting its capability to apprehend intricate patterns within exhaled breath data. Furthermore, observed disparities at baseline between COPD patients who later developed lung cancer and those who did not underscore the model's potential as a prognostic tool. These findings exemplify the capacity of technology-driven diagnostics to reshape the arena of respiratory disease identification. The model's precise differentiation, combined with its elevated sensitivity and specificity, holds potential for early detection and intervention. In this scenario, eNose technology emerges as an innovative instrument in the realm of healthcare, spotlighting the integration of cutting-edge sensor arrays and machine learning algorithms to elevate patient care and prognosis.