COMPUTERIZED ANALYSIS OF INTERSTITIAL INFILTRATES ON CHEST RADIOGRAPHS - A NEW SCHEME BASED ON GEOMETRIC PATTERN FEATURES AND FOURIER-ANALYSIS

Rationale and Objectives. Detection of interstitial infiltrates on che st radiographs is difficult and subjective. Therefore, we developed a computerized method to provide quantitative analysis of lung texture t o increase diagnostic accuracy. Methods. Two hundred chest radiographs -100 healthy and 100 abnormal with interstitial infiltrates-were digit ized using a laser scanner. They were analyzed by an automated compute rized scheme that uses a combination of two methods for detection of i nterstitial infiltrates: a lung texture analysis based on the Fourier transform and a geometric pattern feature analysis based on filtering techniques, Results. The overall sensitivity and specificity of the co mputerized scheme were 92% and 90%, respectively. The scheme achieved a sensitivity of 80% in subtle cases (n = 15) and 88% in cases with lo calized interstitial disease (n = 26), whereas the specificity remaine d unchanged. There was good correlation between the computer output an d the radiologists' severity rating, Conclusion. This enhanced compute rized scheme exhibits high sensitivity and specificity with a large da tabase.