Marius George Linguraru DPhil, MS, Presenter: Nothing to Disclose

Shijun Wang, Abstract Co-Author: Nothing to Disclose

Furhawn Shah, Abstract Co-Author: Nothing to Disclose

Rabindra Gautam, Abstract Co-Author: Nothing to Disclose

W. Marston Linehan MD, Abstract Co-Author: Nothing to Disclose

Ronald M. Summers MD, PhD, Abstract Co-Author: Royalties, iCAD, Inc, Nashua, NH Grant, iCAD, Inc, Nashua, NH Stockholder, Johnson & Johnson

James Peterson, Abstract Co-Author: Nothing to Disclose

00030490-DMT et al, Abstract Co-Author: Nothing to Disclose

To investigate the potential of computer-aided analysis of contrast-enhanced CT data to quantify and classify kidney lesions.

A computer-assisted radiology tool based on level sets was employed to segment 116 renal lesions from 40 contrast-enhanced three-phase abdominal CT studies. Lesion diameter varied from 5.3 to 43.3 mm. There were 41 cysts and 75 cancers: 22 Von Hippel-Lindau (VHL) syndromes, 13 Birt-Hogg-Dube (BHD) syndromes, 19 hereditary papillary renal cell (HPRC) carcinomas, and 21 hereditary leiomyomatosis and renal cell cancers (HLRCC). The technique allowed quantifying the three-dimensional size and volume of lesions. Intra-patient and inter-phase registration facilitated the study of lesion enhancement. The histograms of curvature-related features (HCF) were used to classify the lesion types via random sampling. The areas under the curve (AUC) were calculated for each receiver operating characteristic (ROC) curve.

Tumors were robustly segmented and the volume overlap between manual and semi-automated quantifications was of 0.8+/-0.05, within the inter-observer variability (0.8+/-0.06). The classification based on lesion appearance, enhancement and morphology between cysts and cancers showed AUC=0.99; for BHD+VHL (solid cancers) vs. HPRC+HLRCC AUC=0.99; for VHL vs. BHD AUC=0.61; and for HPRC vs. HLRCC AUC=0.9. All classifications were statistically significant (p<0.05). The Pearson correlation coefficients between the linear/volumetric clinical manual measurements and the semi-automatic quantifications were 0.98/0.99. There was no significant difference between the clinical and semi-automatic linear/volumetric measurements (p>0.05/0.2).

The clinical tool allows the accurate quantification of cystic, solid and mixed renal tumors. Cancer types are further classified into four categories with statistical significance. Computer-assisted image analysis shows great potential for tumor diagnoses and monitoring.

Computer-aided analysis of renal neoplasms on abdominal CT may facilitate their noninvasive classification, guide clinical management, and monitor responses to drugs or interventions.

Linguraru, M, Wang, S, Shah, F, Gautam, R, Linehan, W, Summers, R, Peterson, J, et al, 0, Computer-aided Clinical Tool for Renal Cancer Quantification and Classification.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8002632.html