Delineation of Lung Tumor with Neural Networks

Abstract

Due to the picture similar region of interest, local area, and the effect of respiration, separating a lung tumor from neighboring tissue from a collection of magnetic resonance images (MRI) faces a number of challenges. However, precise tumor segmentation is needed when preparing radiation therapy in order to prevent excessive radiation exposure to healthy tissues. Delineation of the whole MRI pattern by hand is boring, time-consuming, and costly. Using neural networks, this research studies the automated tracking of tumor borders during radiation therapy. To improve the accuracy of lung tests, we proposed using neural network architecture with fuzzy clustering.