ABSTRACT
PURPOSE: Despite clinically available high-resolution CT, the detection and classification of gallstones remains a challenge in some cases. This pilot study examines whether noninvasive characterization of gallstones in vitro is possible using dual-energy analysis (DECT) of dual source CT datasets. MATERIALS AND METHODS: A total of 43 gallstones (0.4 - 1.5 cm) were examined at 80 kV, 140 kV and in the dual-energy mode. The monoenergetic datasets were examined by two independent examiners and classified as calcium, cholesterol or pigment stones. The results were compared with the pathological classification as the clinical gold standard. After creating reference images for each group via dual-energy analysis, the classification was repeated and compared with the gold standard again. RESULTS: Using the monoenergetic analysis at 80 kV, the sensitivity and specificity were 100 / 84 % and 100 / 88 % for calcium stones. For cholesterol stones the values were 54 / 89 % and 54 / 85 % and for pigment stones 70 / 80 % for both examiners. At 140 kV, the sensitivity and specificity for calcium stones were 100 / 84 % for both examiners, 46 / 92 % for cholesterol stones for both examiners and the sensitivity and specificity were 80 / 75 % and 80 / 80 % for pigment stones. Using the reference images established by DECT, both examiners were able to correctly classify all gallstones. CONCLUSION: The present data indicates that DECT is able to correctly classify Gallstones according to the clinical gold standard in vitro. Clinical studies have to demonstrate whether these results lead to optimized clinical decision making.