Patient-specific contrast injection protocols for cardiovascular multidetector row computed tomography

OBJECTIVE: To develop patient-specific contrast injections for uniform enhancement of cardiovascular multidetector row computed tomography (MDCT) images. METHODS: Sixty-two patients were imaged using electrocardiogram (ECG)-gated spiral MDCT. Thirty patients (group 1) received a uniphasic injection; the remaining 32 patients (group 2) received patient-specific multiphasic injections. For group 2 patients, the vasculature between injection and imaging sites was considered a "gray box" whose transfer function was determined from a test bolus injection and the resulting enhancement in the left side of the heart. This transfer function was used to determine the injection necessary to achieve 250 Hounsfield units in the left side of the heart. Intraindividual and interindividual variation of enhancement were determined for both groups. Superior vena cava (SVC) artifacts were graded on a 4-point scale. RESULTS: The measured indices of intraindividual variation were significantly smaller in group 2 than in group 1 (P < 0.05), indicating improved uniformity with patient-specific injections. The interindividual variation of mean enhancement in group 2 was smaller than in group 1, but the difference was not significant. The severity of SVC artifacts was significantly reduced (P < 0.05) for thinner patients (<83 kg) in group 2 compared with similar patients in group 1. CONCLUSIONS: Patient-specific multiphasic contrast injections yielded more uniform enhancement in the left side of the heart on MDCT images with reduced intraindividual variation of enhancement compared with standard uniphasic injections. Patient-specific injections also reduced SVC artifacts in patients <83 kg.