An IoT Framework for the Detection of Lung Cancer Using a Decision Support System
Ahamd Khader Habboush, Bassam Mohammad Elzaghmouri, Binod Kumar Pattanayak, Pravat Kumar Rautaray
Abstract
Cancer remains an ongoing global health challenge, necessitating the progress of innovative techniques for early detection and risk assessment. In this study, a comprehensive method is presented for predicting lung cancer by utilizing a carefully curated dataset consisting of 1000 individuals from the Kaggle dataset. Cutting-edge machine learning models were leveraged, including Support Vector Machines (SVM), Naïve Bayes Multinomial (NBM), KNN, PART (Partial Rule-based Tree), and Random Forest (RF), to improve the precision of our forecasts. The dataset that was compiled included a wide array of patient characteristics, encompassing demographics, lifestyle factors, medical history, and health data gathered through IoT devices. By harnessing the capabilities of IoT technology, real-time and continuous health monitoring was enabled, facilitating a dynamic assessment of lung cancer risk. The findings revealed that the PART model achieved an impressive accuracy of 91%, surpassing other models like Random Forest (80%), K-Nearest Neighbors (83%), SVM (89%), and Naïve Bayes Multinomial (86%). This innovative approach shows promise in the early detection of lung cancer and the provision of personalized risk assessments, possibly resulting in better patient outcomes and decreased healthcare challenges.