Automated staging of coal workers' pneumoconiosis utilizing digital image analysis of chest x-rays.
Preston-K Jr.; Witt-EK
Electrical and Computer Engineering Department and Biomedical Engineering Program, Carnegie Mellon University, Pittsburgh, Pennsylvania 1996 Feb; :1-15
This final report covers research designed to determine the feasibility of automated staging of coal workers' pneumoconiosis (CWP) using digital image analysis of chest X-rays. A system which can potentially aid the radiologist in the classification of CWP from chest X-rays is described. The system applies statistical pattern recognition to digitized CWP chest X-ray images based on the generation of features from the matrix of numbers that represents the digitized films, and the application of a variety of algorithms that may be used to distinguish profusion and opacity size. Both traditional (2-D Fourier transform and the spatial gray level dependence matrix) and new (minimum tracking algorithm and the Gauss- Markov random field parameters) feature sets were used; 386 features were included. Comparison of an expert radiologist's classifications with those of other radiologists yielded correlation coefficients of 0.60 and 0.82 for profusion and opacity size, respectively. When the expert's classifications were compared to the computer classifications, the correlation coefficients were 0.56 for profusion and 0.55 for opacity size. The authors conclude that ultimately the classifier and visualization tools described can be used in a preliminary chest X-ray screening system which would identify regions possibly exhibiting pathology, and indicate these regions to the radiologist.
NIOSH-Grant; Pulmonary-system-disorders; Diagnostic-techniques; Coal-workers-pneumoconiosis; Respiratory-system-disorders; Lung-disease; Occupational-respiratory-disease
Electrical Engineering Carnegie Mellon University 5000 Forbes Ave Pittsburgh, PA 15213
Final Grant Report
NTIS Accession No.
Electrical and Computer Engineering Department and Biomedical Engineering Program, Carnegie Mellon University, Pittsburgh, Pennsylvania
Carnegie-Mellon University, Pittsburgh, Pennsylvania