Ce. Floyd et al., DIFFUSE NODULAR LUNG-DISEASE ON CHEST RADIOGRAPHS - A PILOT-STUDY OF CHARACTERIZATION BY FRACTAL DIMENSION, American journal of roentgenology, 167(5), 1996, pp. 1185-1187
OBJECTIVE. We present a computer-aided diagnostic technique for identi
fying nodular interstitial lung disease on chest radiographs. The frac
tal dimension was used as a numerical measure of image texture on digi
tal chest radiographs to distinguish patients with normal lung from th
ose with a diffuse nodular interstitial abnormality. MATERIALS AND MET
HODS. Twenty digitized chest radiographs were classified as normal (n
= 10) or as containing diffuse nodular abnormality (n = 10) on the bas
is of readings assigned according to the classification of the Interna
tional Labour Organization. Regions of interest (ROIs) measuring 1.28
cm(2) were selected from the intercostal spaces of these radiographs.
The fractal dimension of these ROIs was estimated by power spectrum an
alysis. The cases were not subtle. RESULTS. The fractal dimension prov
ided statistically significant discrimination between normal parenchym
a and nodular interstitial lung disease. The area under the receiver o
perating characteristic curve was 0.90 (+/-0.02). One operating point
provides sensitivity of 88% with a specificity of 80%. CONCLUSION. The
fractal dimension can provide a measure of lung parenchymal texture a
nd shows promise as an element of computer-aided diagnosis, characteri
zation, and follow-up of interstitial lung disease.