Clinical Informatics System to Understand Lung Cancer Using Transfer Learning-Based Computer Aided Application-I

Significant advancements in diagnostic and sequencing technologies have propelled lung cancer clinical studies forward. The focal point of these endeavors has been the early detection of lung cancer, leading to the development of various computer-aided diagnostic tools powered by Artificial Intelligence applications and deep learning techniques. However, it is note- worthy that some deep learning techniques are susceptible to overfitting, which can compromise their effectiveness in practice. Addressing the challenge of accurately categorizing lung cancer is crucial. In this study, we introduced a CNN-based classification method designed to distinguish between lung adenocarcinoma and benign lung conditions. To enhance model performance, we incorporated batch normalization into our proposed approach, which yielded impressive results across both binary categories, achieving an accuracy rating of 97.92%. In our comparative analysis with other deep learning-based methods, our approach consistently demonstrated superiority in performance. This underscores the potential of our model as a valuable tool in the accurate categorization of lung cancer.