Akademik Veri Yönetim Sistemi
14th International Conference on Electrical and Electronics Engineering, ELECO 2023, Virtual, Bursa, Türkiye, 30 Kasım - 02 Aralık 2023
Lung cancer remains a major global health problem that requires early and accurate diagnosis to improve patient outcomes. Assessment by specialists is time-consuming, tedious and leads to diagnostic inconsistencies. This has led to the development of computer-aided diagnosis systems for lung cancer diagnosis. Traditionally, proposed systems have used a transfer learning approach with CNN-based models. It has also been observed that cross-validation is usually not applied in these studies. On the other hand, cross-validation contributes to more reliable results by enabling the developed models to generalise to various image samples. In recent years, Vision Transformer (ViT) models have shown remarkable success in various computer vision tasks, including image classification. In this paper, we present our pioneering approach to lung cancer classification using various Vision Transformer-based models. However, we recognised the potential of these models to capture fine-grained patterns in medical images. To ensure the robustness and reliability of our models, we also introduced 5-fold cross-validation as a key component of our methodology. The proposed system yielded the highest results with 99.69% accuracy, 99.62% precision and 98.80% recall. This research represents an important step in utilising the latest computer vision techniques to improve lung cancer diagnosis, ultimately contributing to better patient care and outcomes.