OBJECTIVE: To assess the capability of the density histogram of high-resolution CT (HRCT) in differentiating nonspecific interstitial pneumonia (NSIP) from usual interstitial pneumonia (UIP) without definite honeycombing. METHODS: Twenty-eight NSIP and 32 UIP patients without definite honeycombing on CT were included in this study. We evaluated a CT image at the level of 1 cm above the diaphragm. The pixels of the lung parenchyma were classified into 4 ranges. The fractions of each range of corresponding density were calculated. The skewness (the degree of asymmetry of a distribution) and kurtosis (how sharply peaked a histogram is) were obtained from the density histogram. The mean value and the mode value of the lung attenuation were also measured. RESULTS: The fraction of the range of ground-glass opacity and reticular opacity was greater in NSIP patients (32%, 12%) than in UIP patients (23%, 8%) (P < 0.001). UIP had a larger fraction of the range of normal lung. The density histogram was less skewed (P = 0.01) and had a wider peak (P = 0.02) in NSIP (skewness = 1.2879 +/- 0.5672, kurtosis = 1.2115 +/- 1.9470) than in UIP (skewness = 1.6426 +/- 0.4664, kurtosis = 2.3880 +/- 1.8183). CONCLUSION: The density histogram reflected the differences in the CT features between NSIP and UIP. Therefore, a density histogram may be helpful for differentiating NSIP from UIP without definite honeycombing.
OBJECTIVE: To assess the capability of the density histogram of high-resolution CT (HRCT) in differentiating nonspecific interstitial pneumonia (NSIP) from usual interstitial pneumonia (UIP) without definite honeycombing. METHODS: Twenty-eight NSIP and 32 UIP patients without definite honeycombing on CT were included in this study. We evaluated a CT image at the level of 1 cm above the diaphragm. The pixels of the lung parenchyma were classified into 4 ranges. The fractions of each range of corresponding density were calculated. The skewness (the degree of asymmetry of a distribution) and kurtosis (how sharply peaked a histogram is) were obtained from the density histogram. The mean value and the mode value of the lung attenuation were also measured. RESULTS: The fraction of the range of ground-glass opacity and reticular opacity was greater in NSIP patients (32%, 12%) than in UIP patients (23%, 8%) (P < 0.001). UIP had a larger fraction of the range of normal lung. The density histogram was less skewed (P = 0.01) and had a wider peak (P = 0.02) in NSIP (skewness = 1.2879 +/- 0.5672, kurtosis = 1.2115 +/- 1.9470) than in UIP (skewness = 1.6426 +/- 0.4664, kurtosis = 2.3880 +/- 1.8183). CONCLUSION: The density histogram reflected the differences in the CT features between NSIP and UIP. Therefore, a density histogram may be helpful for differentiating NSIP from UIP without definite honeycombing.
Authors: David J Lederer; Paul L Enright; Steven M Kawut; Eric A Hoffman; Gary Hunninghake; Edwin J R van Beek; John H M Austin; Rui Jiang; Gina S Lovasi; R Graham Barr Journal: Am J Respir Crit Care Med Date: 2009-06-19 Impact factor: 21.405
Authors: N Sverzellati; E Calabrò; A Chetta; G Concari; A R Larici; M Mereu; R Cobelli; M De Filippo; M Zompatori Journal: Radiol Med Date: 2007-12-13 Impact factor: 3.469
Authors: Nicholas D Weatherley; James A Eaden; Neil J Stewart; Brian J Bartholmai; Andrew J Swift; Stephen Mark Bianchi; Jim M Wild Journal: Thorax Date: 2019-03-18 Impact factor: 9.139
Authors: Bina Choi; Steven M Kawut; Ganesh Raghu; Eric Hoffman; Russell Tracy; Purnema Madahar; Elana J Bernstein; R Graham Barr; David J Lederer; Anna Podolanczuk Journal: ERJ Open Res Date: 2020-03-02