Lung cancer, a dangerous pulmonary tumor, significantly impacts an individual’s health. In the lack of automated and clear diagnostic technologies, doctors use biopsy as the final test. However, biopsy can be painful and expensive. Diagnostic mistakes and dataset accessibility restrictions also challenge researchers. This paper presents a model that initially categorizes a CT scan image as Normal, Benign, or Malignant, then subsequently calculates the tumor percentage present. The lung cancer dataset from the Iraq-Oncology Teaching Hospital/National Center for Cancer Diseases (IQ-OTH/NCCD) has been utilized. The principal objective of this paper is the early diagnosis of lung cancer through the evaluation of classification algorithms, thereby significantly alleviating the workload of physicians to enable timely medical intervention for numerous lung cancer patients. Additionally, it is demonstrated how genomic data can be reshaped for potential use in similar machine learning methods, offering a pathway to validate and enhance the robustness of image-based classification results.