Lung Nodule Classification Using Scale-Invariant Neural Networks

The size of a lung nodule is very diverse. The high variation of nodule sizes makes classifying them a difficult and challenging task. In this thesis, we propose four different Scale-Invariant Neural Networks for lung nodule classification. We examined our proposed methods on two different public datasets and compared their performance with state-of-the-art literature methods. The results show that the proposed model outperforms state-of-the-art methods and achieves an AUC of 98.05% and accuracy of 95.28% on the LIDC-IDRI dataset. These results will lead to identifying lung cancer in its early stages.