Improved detectability of thromboses of the lower limb using low kilovoltage computed tomography

To determine the utility of low kilovoltage computed tomographic venography (CTV) for the detection of deep venous thrombus in the lower limbs. Twenty-one thrombi in 19 enrolled patients were investigated in this retrospective study. Patients were initially scanned using CTV at 100kVp, at the femur level, followed by an immediate switch to 80kVp. We assessed the CT values of thrombi and veins and performed subjective evaluation for detecting thrombi using a 5-point scoring system: 1, unable to evaluate due to noise or artifacts; 2, equivocal venous thrombus; 3, possible venous thrombus; 4, probable venous thrombus; and 5, definite venous thrombus. Venous density on 100-kVp images (mean±SD [standard deviation]: 122±23HU, 95% confidence interval [CI]: 111-133 Hounsfield unit [HU]) was significantly lower than that on 80-kVp images (136±24HU, 95% CI: 125-147HU, P<.001). There was no significant difference in thrombi between 100-kVp images (55±14HU, 95% CI: 49-61HU) and 80-kVp images (57±16, 95% CI: 50- 64HU, P=.168). The thrombus to vein ratio on 100-kVp images (0.47±0.20, 95% CI: 0.39-0.56) was significantly higher than that on 80-kVp images (0.44±0.16, 95% CI: 0.37-0.51, P=.048). The mean 5-point score was significantly higher on the 80-kVp images (4.76) than on the 100-kVp images (4.45, P=.016). Lower kilovoltage CTV significantly improved thrombotic to venous contrasts in the lower limbs.