Optimizing CT pulmonary angiography: Manual versus automatic kVp selection

INTRODUCTION: Automatic selection of CT scanning parameters is widely used in current clinical practice, but these systems rely on algorithmic assumptions that may not hold for all patients or indications. Aim of our study was to evaluate whether manual peak tube voltage (kVp) selection can outperform a fully automatic kVp selection algorithm in CT pulmonary angiography (CTPA) by reducing radiation dose without compromising image quality. METHODS: A total of 2569 CTPA scans were performed in an emergency setting using a 256-slice dual-source CT scanner. Baseline imaging (n = 750) employed full dose modulation with automatic kVp selection. In the optimized protocol (n = 1819), radiographers manually selected a fixed kVp (80 or 100) based on the topogram and dose modulation curve. Radiation dose metrics were extracted using Radimetrics (Bayer). Image quality was assessed in a subset of 15 patients scanned with both protocols by measuring contrast-to-noise ratio (CNR) and CNR per dose (CNRD) in the pulmonary arteries. RESULTS: Automatic kVp selection predominantly assigned 120 kVp, accounting for 61.5% of patients in that group. In contrast, manual selection favored 80 kVp, used in 66.2% of patients. The optimized protocol significantly reduced effective dose (3.6 ± 1.8 mSv vs. 1.9 ± 0.8 mSv, p < 0.01). CNR was comparable between the automatic and manual protocols (18.6 ± 3.9 vs. 22.3 ± 3.5, p = 0.12), while CNRD improved significantly with manual selection (8.1 ± 3.0 vs. 12.6 ± 5.5, p < 0.01). CONCLUSIONS: Optimized manual selection of kVp can substantially reduce patient radiation dose in CTPA without degrading diagnostic image quality, outperforming a fully automatic kVp selection algorithm. IMPLICATIONS FOR PRACTICE: Manual kVp selection offers a practical way to reduce CTPA dose in fast-paced emergency imaging environments.