engleski

Histogram Analysis of X-ray Images and Wavelet Influence to the Contained Information

We used ILO (International Labor Organization) classification and real-patients' X-ray pulmonary images to investigate the possibility of computer-aided diagnosis of the asbestosis. We used digitized pulmonary images, in tiff (tagged interchange format) and bitmap format, and Matlab to determine whether patient is healthy or not. In our limited research, our approach gave 100% correct answers, but further research is needed, as well as the further development of the approach. Our approach gives opportunity to speed up diagnostics procedure, improve health service and reduce costs and loss of working hours. Statistical analysis of the pulmonary X-ray image' s gray tones is the key of the approach and gives an opportunity to the computer to evaluate the patient' s condition objectively. Furthermore, we used Daubechies wavelets to test the approach' s resistance to wavelet compression, and concluded that wavelet transform can be used for these purposes only, for other purposes, the effect can only be speculated on and the WT use must be approached carefully.

wavelets; asbestosis; medical imaging

Medical and Biological Engineering and Computing ; 37 (1999) S2