An algorithm for feature extraction and detection of pulmonary nodules in digital radiographic images
Cesar Supanta, Guillermo Kemper, Christian del Carpio · 2018 IEEE International Conference on Automation/XXIII Congress of the Chilean Association of Automatic Control (ICA-ACCA) · 2018
This work proposes a method for feature extraction and detection of pulmonary nodules in digital radiographic images, as little visualization and highlighting of these features often prevent a deeper diagnosis in chest radiographs. The proposed method involves digital image processing techniques such as re-quantization, gamma correction, OTSU thresholding, projection analysis, convergence filter, dilation, erosion and geometric filters. The proposed algorithm has a sensitivity of 91%, specificity of 96% and precision 94% with a referential database of 50 chest radiographs.