Attention Based Recurrent Neural Network for Lung Cancer Detection

Lung cancer detection is a significant technique needed in the current medical field. Moreover, it causes the larger death rate because of late identification which then delays the diagnosis process. Hence, to provide the earlier detection of lung cancer, Attention-based recurrent neural network (ARNN) is designed in our work. This proposed approach is framed for the motive of obtaining the increased accuracy and faster prediction of the cancer spots even before the cancer symptoms recognized by the patients. The proposed ARNN has an attention layer which performs the encoding and decoding process within single sequences using variable-length vector and also has an advantage that it performs the automatic feature extraction. Finally, the classification is done using the proposed ARNN and the classification results are obtained with great accuracy and within less computation time.