Hybrid Intelligent Models for Chest X-Ray Image Segmentation

Abstract

Hybrid intelligent systems allowing segments related to diagnosable diseases to be identified on chest X-ray films are presented. X-Rays were segmented by analysis of two-dimensional Fourier spectra with a sliding window. The spectrum of the sliding window was processed with sequential filters constructed on the basis of different image processing paradigms. An algorithm for tuning the filters of hybrid models is proposed, and results from segmentation of X-rays from patients with pneumonia are demonstrated.