Comparison of machine learning methods for classifying mediastinal lymph node metastasis of non-small cell lung cancer from 18F-FDG PET/CT images

摘要

The present study shows that the performance of CNN is not significantlydifferent from the best classical methods and human doctors for classifyingmediastinal lymph node metastasis of NSCLC from PET/CT images. Because CNN doesnot need tumor segmentation or feature calculation, it is more convenient andmore objective than the classical methods. However, CNN does not make use ofthe import diagnostic features, which have been proved more discriminative thanthe texture features for classifying small-sized lymph nodes. Therefore,incorporating the diagnostic features into CNN is a promising direction forfuture research.