Protocol modifications for CT perfusion (CTp) examinations of abdomen-pelvic tumors: Impact on radiation dose and data processing time

Purpose: To evaluate the effect of CT perfusion (CTp) protocol modifications on quantitative perfusion parameters, radiation dose and data processing time. Materials & methods: CTp datasets of 30 patients (21M:9F) with rectal (n=24) or retroperitoneal (n=6) tumours were studied. Standard CTp protocol included 50 sec cine-phase (0.5 sec/rotation) and delayed-phase after 70 ml contrast bolus at 5-7 ml/sec. CTp-data was sub-sampled to generate modified datasets (n=105) with cine-phase(n=15) alone, varying cine-phase duration (20-40 sec, n=45) and varying temporal sampling-interval (1-3 sec, n=45). The estimated CTp parameters (BF, BV, MTT&PS) and radiation dose of standard CTp served as reference for comparison. Results: CTp with 50 sec cine-phase showed moderate to high correlation with standard CTp for BF&MTT (r=0.96&0.85) and low correlation for BV (0.75, p=0.04). Limiting cinephase duration to 30 sec demonstrated comparable results for BF&MTT, while considerable variation in CTp values existed at 20 sec. There was moderate-to-high correlation of CTp parameters with sampling interval of 1&2 sec (r=0.83-0.97, p>0.05), while at 3 sec only BF showed high correlation (r=0.96, p=0.05). Increasing sampling interval (47-60%) and reducing cine-phase duration substantially reduced dose(30.8-65%) which paralleled reduced data processing time (3-10 min). Conclusion: Limiting CTp cine-phase to 30 sec results in comparable BF&MTT values and increasing cine-phase sampling interval to 2 sec provides good correlation for all CTp parameters with substantial dose reduction and improved computational efficiency.