Determination of renal stone composition in phantom and patients using single-source dual-energy computed tomography

This study aimed to characterize the urinary tract stones in phantom and patients using single-source dual-energy computed tomography. Twenty stones of pure crystalline composition (uric acid [UA], struvite, cystine, and calcium oxalate monohydrate) were assessed in a phantom and 11 patients (age 39-67 years) with urinary tract stones were evaluated. An initial low-dose unenhanced CT (tube potential, 120 kilovolts [peak]; milliampere range, 150-450; noise index, 26; section thickness, 5 mm) followed by a targeted dual-energy computed tomography acquisition on a single-source dual-energy computed tomography (Discovery CT 750 HDCT, GE) was performed. Uric acid and non-UA stones were defined using a 2-material decomposition (material density-iodine/water) algorithm. The stone effective atomic number (Zeff) was used to subclassify non-UA stones. The stone attenuation (Hounsfield unit) was also studied to determine its performance in predicting the composition. Ex vivo chemical analysis of the stone served as a criterion standard. Of the 59 verified stones (phantom, 20; patients, 39; mean size, 6 mm), there were 16 UA and 43 non-UA type. The material density images were 100% sensitive and accurate in detecting UA and non-UA stones. The Zeff accurately stratified struvite, cystine, and calcium (calcium oxalate monohydrate) stones in the phantom. In patients, Zeff identified 83% of calcium stones (n = 24), and in stones of mixed type, it resembled dominant composition. The Hounsfield unit measurements alone were 71% sensitive and 69% accurate in detecting the UA stones. Single-source dual-energy computed tomography can accurately predict UA and non-UA stone composition in vitro and in vivo. Substratification of non-UA stones of pure composition can be made in vitro and in vivo. In stones of mixed composition, the Zeff values reflect the dominant composition.