Image feature analysis and computer-aided diagnosis in digital radiography: detection and characterization of interstitial lung disease in digital chest radiographs

Abstract

We are developing an automated method for determining physical measures of lung textures in digital chest radiographs in order to detect and characterize interstitial lung disease. With this method, the underlying background density variations caused by the gross lung and chest wall anatomy are corrected for in order to isolate the fluctuating patterns of the underlying lung texture for subsequent computer analysis. The power spectrum of lung texture, which is obtained from the two-dimensional Fourier transform, is filtered by the visual system response of the human observer. The magnitude and coarseness (or fineness) of the lung textures are then quantified by the root-mean-square (rms) variation and the first moment of the power spectrum, respectively. Preliminary results indicate that the rms variations and/or the first moments of the texture of abnormal lungs with various interstitial diseases are clearly different from those of normal lungs. Our results suggest strongly that quantitative texture measures calculated from digital chest images may be useful to radiologists in their assessment of interstitial disease.