Value of three-dimensional spiral CT hepatic angiography

The aim of this study was to compare with conventional angiography images the diagnostic utility of three-dimensional CT hepatic angiography (CTHA) images reconstructed from spiral CT images obtained during direct injection of contrast material into hepatic arteries. Three-dimensional CTHA images were reconstructed from spiral CTHA images obtained during direct hepatic arterial contrast material injection in 22 patients. We used maximum-intensity-projection (MIP) and shaded-surface-display (SSD) techniques. We compared matched pairs of MIP CTHA images and SSD CTHA images with conventional angiograms for depiction of peripheral (segmental to subsegmental order branches) and proximal (common to proper) hepatic arteries as well as for conspicuity of the entire hepatic arterial tree in 22 patients and for conspicuity of tumors, recognition of tumor sites, and depiction of tumor-feeding arteries in 21 patients with malignant liver tumors. Our evaluations were performed in a blinded fashion by two experienced radiologists, who reached a consensus. (These radiologists also counted the number of tumors depicted by each method in the 21 patients with malignant liver tumors.) Depiction of peripheral and proximal hepatic arteries and conspicuity of the entire hepatic arterial tree were significantly better on conventional angiograms than on MIP CTHA images and SSD CTHA images. Conspicuity of tumors and recognition of tumor sites were significantly better on MIP CTHA images and SSD CTHA images than on conventional angiograms. Depiction of tumor-feeding arteries was significantly better on MIP CTHA images than on SSD CTHA images and conventional angiograms. For 47 malignant liver tumors, MIP CTHA images showed 40 (85%), SSD CTHA images showed 34 (72%), and conventional angiograms showed 20 (43%). MIP CTHA images showed liver tumors, tumor sites, and tumor-feeding arteries better than conventional angiograms. MIP was the preferred method for three-dimensional CTHA analysis.