Stone characteristics on noncontrast computed tomography: establishing definitive patterns to discriminate calcium and uric acid compositions.

To evaluate if different stone compositions can be distinguished by variations in Hounsfield unit (HU) patterns from stone core to periphery. We retrospectively reviewed our stone registry searching for pure uric acid (UA) and calcium oxalate (CaOx) stones. Inclusion criteria comprised: pure calculi; noncontrast computed tomography before stone analysis; and stone size ≥4 mm. A single urologist reviewed all images (abdominal/bone windows). Absolute HU attenuation was measured in the center (core) and edges (periphery) of each stone. HU density (HU/size) and stone volume were calculated. Categorical variables were compared with Chi-square/Fisher exact test; continuous with analysis of variance/Student t test. Spearman's test was used correlate HU values and stone dimensions. A P <.05 was considered significant. Included were 47 UA and 36 CaOx stones. CaOx cohort had higher HU values for all parameters (P <.001). HU variation (core-periphery) was more homogeneous in the UA group (-25%) than in the CaOx group (-52%). Stone dimensions had a significant positive correlation with HU values in CaOx cohort. In the UA group, stone volume and periphery HU had significant positive correlation. On bone window, CaOx stones had a heterogeneous pattern with a higher attenuation core with decreasing attenuation toward the periphery; UA calculi had a more homogeneous constant frosted white cloudy appearance. Stone size does not affect HU for UA stone, but has a positive correlation with CaOx stone. UA stones can be differentiated from CaOx by evaluating variation in HU from core to periphery and by subjective evaluation of homogeneity in bone windows.