Katharina Marten MD, Abstract Co-Author: Nothing to Disclose

Volker Dicken, Abstract Co-Author: Nothing to Disclose

Christian Kneitz, Abstract Co-Author: Nothing to Disclose

Werner Kenn MD, Presenter: Nothing to Disclose

Margarete Hoehmann, Abstract Co-Author: Nothing to Disclose

Dietbert Hahn MD, Abstract Co-Author: Nothing to Disclose

Christoph Engelke MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

To evaluate a threshold-based computer-aided diagnosis (CAD) tool (MeVisPULMO 3D) for quantification of interstitial lung disease (ILD) in patients with collagen vascular disease.

50 patients with rheumatoid arthritis (RA, n=24), scleroderma (PSS, n=14) and systemic lupus erythematosus (SLE, n=14) were included and underwent thin-section CT (16x.75mm collimation). CT scans were assessed by two observers for extent of ILD (EoILD), extent of reticulation (EoRet) and extent of ground glass opacity (EoGGO), and twice by CAD for EoILD. Pulmonary function tests were obtained in all patients. Statistical evaluation used 95% limits of agreement (LoA) and linear regression analysis.

There was total concordance between the CAD measurements. Interobserver agreement was good (LoA=-27.0 to 17.0% and -24.5 to 15.5%, for EoILD and EoRet, respectively). EoILD by readers correlated strongly with DLCO (r=-.705, p<.0001) and moderately with FVC (r=-.559, p=.0002). EoRet correlated closely with DLCO and moderately with FVC (DLCO: r=-.663; FVC: r=-.436; p≤.005). There was close correlation between readers and CAD (EoILD vs CAD: r = .716, p<.0001; EoRet vs. CAD: r=.69, p<.0001). CAD correlated well with DLCO (r=-.531, p<.0001) and moderately with FVC (r=-.483, p=.0008). Subgroup analysis (EoGGO <15%) strengthened the correlations between readers and CAD, readers and PFT, and between CAD and PFT. Stratification for patient diagnosis had no effect (P>.05) on any correlation.

The CAD system used holds promise to become a valuable tool for quantification of ILD, showing close correlation with human observers and physiologic impairment.

The CAD application used proved to be time-efficient, robust and reproducible, and holds promise to become a valuable tool for quantification of interstitial lung disease.

Marten, K, Dicken, V, Kneitz, C, Kenn, W, Hoehmann, M, Hahn, D, Engelke, C, et al, , Interstitial Lung Disease Associated with Collagen Vascular Disorders: Disease Quantification Using a Computer-aided Diagnosis Tool.  Radiological Society of North America 2008 Scientific Assembly and Annual Meeting, February 18 - February 20, 2008 ,Chicago IL. http://archive.rsna.org/2008/6007250.html