An efficient and complete automatic system for detecting of lung module
S. Balakrishnan, J. Janet, K. Sujatha, S. Sheeba Rani · Indian Journal of Science and Technology · 2018
Objectives: To make a fully automated algorithm that is based on simple and quick steps, which produces consistent output for the same inputs. Methods/Statistical Analysis: For thorax and lung segmentation, region growing based method is used to segment the region of interest. The missing parts of the lungs are reconstructed using morphological operations. After that, nodules are detected based on the features of the reconstructed image. Artificial Neural Network has been used for classifying the images. Findings: An aggregate of 100 pictures with determination of 512 × 512 pixels with eight bits for every shading channel are caught. 90% affectability was obtained with 0.05 false positives for each picture. Application/Improvements: This framework distinguishes the phase of lung malignancy. The outcomes demonstrate that the tumors are of various measurements. By estimating the measurements of the tumor the lung disease stage can be recognized precisely utilizing the proposed technique. The outcomes indicate great potential for lung growth identification at beginning time. Keywords: Automatic System, Filter, Lung Cancer, Neural Network, Region Based Method