Lung cancer remains one of the most common causes of cancer mortality globally, an‎d early detection is important to improve the survival of patients. Chest X-ray is a commonly used diagnostic imaging technique that is inexpensive an‎d non-invasive, but interpreting the images by han‎d is challenging because the difference in appearance between normal an‎d malignant tissues is subtle. The overall goal of this project is to automate the detection of lung tumors in chest X-ray images using machine learning to increase overall accuracy an‎d efficiency. The dataset consisted of labeled X-ray images of the chest categorized into normal an‎d tumor classes. Three machine learning approaches—Ran‎dom Forest (RF), Support Vector Machine with RBF kernel (SVM-RBF), an‎d Extreme Gradient Boosting (XGBoost)—were developed an‎d eva‎luated to determine the best machine learning model for classifying images. The models are derived using preprocessed image data, extracting features before the starting classification an‎d eva‎luation. The identified metrics illustrate the performance of the model, using metrics of accuracy, precision, recall, F1-score, an‎d confusion matrix plots. The first run of the experiments yield promising results, with average classification accuracy of over 77%, an‎d histograms are eva‎luated for optimization for performance improvements to exceed an accuracy of 90%. The generated trained final models were exported for operational use in a web application to facilitate diagnostic use an‎d deployable for real world assistance. The work illustrates an‎d demonstrates the benefit of conventional machine learning algorithms used alongside medical imaging to assist radiologists with lung cancer detection an‎d mitigate the chance of misdiagnosis.

10/27/2025 12:00:00 AM

87901

1404/08/06

Study on Lung Cancer Diagnosis using Data Mining an‎d AI Methods

(تشخيص سرطان ريه، طبقه‌بندي اشعه ايكس قفسه سينه، يادگيري ماشين، تحليل تصاوير پزشكي، طبقه‌بندي تومور، تشخيص به كمك كامپيوتر (CAD))

(Lung Cancer Detection, Chest X-Ray Classification, Machine Learning, Medical Image Analysis, Tumor Classification, Computer-Aided Diagnosis (CAD))