Lung cancer remains one of the leading causes of cancer-related mortality worldwide. Accurate early detection is essential to improving patient outcomes, yet it remains a challenge, especially in low-resource settings where access to imaging and genetic testing is limited. In this study, we introduce a high-performance, fully interpretable ensemble-based machine learning model that achieves 100% accuracy on the publicly available "Cancer Patients and Air Pollution" dataset. The model uses only clinical and environmental tabular data and significantly outperforms prior work. Extensive validation through stratified cross-validation and baseline comparison confirms that the performance is not due to over fitting or data leakage. The proposed framework has real-world applicability as a non-invasive, low-cost, and deployable tool in clinical screening programs. SHAP-based explain ability analysis further enhances its trustworthiness, paving the way for AI-driven early diagnosis in resource-constrained environments.