Lung cancer is a leading cause of cancer-related deaths globally. Early diagnosis is crucial, but the variability in imaging and histological characteristics makes accurate and timely diagnosis challenging. This systematic review assesses the diagnostic accuracy of machine learning (ML) AI architectures in lung cancer detection and classification. Nine studies using various ML architectures (ANN, EDM, PNN, SVM, POMDP, RFNN) were analyzed. Results showed promising diagnostic accuracy across different lesion types and imaging modalities, although performance varied across studies. The review highlights the potential of AI as a supplementary tool for lung cancer diagnosis while emphasizing the need for further research to optimize AI algorithms and evaluate their performance in diverse populations.

A.C. Pacurari, S. Bhattarai, A. Muhammad, C. Avram, A.O. Mederle, O. Rosca, F. Bratosin, I. Bogdan, R.M. Fericean, M. Biris, G. Tofan, F. Olteanu, A. Mavrea, C. Dima