Intelligent Expert System for Lung Cancer Prediction using Soft Computing

Lung cancer is one of the major causes of death worldwide, and premature diagnosis is the fundament of improving patient outcomes. The project is centred on the development of an expert system for lung cancer risk prediction based on a fuzzy logic-based rule-based inference engine in Java. The system accepts user inputs—age, smoking, chronic cough, chest pain, pollution, and cancer history in the family—and classifies lung cancer risk. Based on fuzzification, inference, and defuzzification, it classifies the diagnosis as low risk, moderate risk, and high risk with a probability score. For simplicity and accessibility, an Android application was developed with real-time prediction and ease of use. The application has an information module for risk factors and preventive care of lung cancer, a predictive module for real-time analysis, and a results module showing risk classification and probability. A case study involved a 62-year-old patient with multiple high-risk factors—heavy smoker (risk 0.9), chronic cough (risk 0.7), chest pain (risk 0.8), high pollution (risk 0.9), and cancer history in the family (risk 0.8). The fuzzy inference engine employed IF-THEN rules to calculate the overall cancer risk of 0.8 (80%) and therefore classify the diagnosis as high risk. This is an illustration of how fuzzy logic improves diagnostic accuracy through consideration of uncertainties in risk levels and symptom severity. With integration of fuzzy logic, real-time decision support, and mobile usability, the system improves early detection, providing individuals and health workers with an effective tool for early medical intervention and preventive healthcare.