Early-Stage Lung Cancer Prediction and Stage Classification Using Machine Learning

Lung cancer is still a major worldwide health issue that necessitates novel strategies for early identification and accurate staging. In response to this critical requirement, we carried out a thorough investigation that makes use of machine learning methods to improve the precision and effectiveness of lung cancer staging and diagnosis, principally through the examination of computed tomography (CT) scan pictures. Our research process includes a number of essential elements. First, to improve picture quality and separate regions of interest, CT scan images underwent extensive image processing using Gabor filtering and watershed segmentation. Then, we carefully extracted features, concentrating on the size and form characteristics of the discovered lung nodules. These characteristics were the basis for developing and testing several machine learning models. With a phenomenal accuracy rate of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 8. 4 4 \%, }$</tex> Convolutional Neural Networks (CNN) stand out as the top performer, according to our data. CNN's uncanny ability to automatically extract complex picture patterns and attributes is credited with its outstanding accuracy, which has revolutionised the prediction of early-stage lung cancer. Along with producing impressive outcomes, K-Nearest Neighbours (KNN), Support Vector Machines (SVM), and Naive Bayes (NB) also showed their adaptability in medical picture analysis. These findings have significant implications for cancer and healthcare. The use of machine learning algorithms offers earlier, more precise diagnosis of lung cancer, opening possibilities to prompt treatments and individualised treatment plans. Additionally, the accurate staging made possible by these models improves the capacity to customise treatments to specific patient demands, thereby enhancing patient outcomes and healthcare procedures.