Dual-energy CT of the pancreas: improved carcinoma-to-pancreas contrast with a noise-optimized monoenergetic reconstruction algorithm.

PURPOSE

To evaluate a novel monoenergetic reconstruction algorithm (nMERA) with improved noise reduction for dual-energy CT (DECT) of pancreatic adenocarcinoma.

MATERIALS AND METHODS

Sixty patients with suspected pancreatic carcinoma underwent dual-source dual-energy CT with arterial phase. Images were reconstructed as linearly-blended 120-kV series (M_0.6) and with the standard monoenergetic (sMERA) and the novel monoenergetic algorithm (nMERA) with photon energies of 40, 55, 70 and 80 keV. Objective image quality was compared regarding image noise, pancreas attenuation, signal-to-noise ratio (SNR) and pancreas-to-lesion contrast. Subjective image quality was assessed by two observers.

RESULTS

Thirty pancreatic adenocarcinomas were detected. nMERA showed significantly reduced image noise at low keV levels compared with sMERA images (55 keV: 7.19 ± 2.75 vs. 20.68 ± 7.01 HU; 40 keV: 7.33 ± 3.20 vs. 37.22 ± 14.66 HU) and M_0.6 (10.69 ± 3.57 HU). nMERA pancreatic SNR was significantly superior to standard monoenergetic at 40 (47.02 ± 23.41 vs. 9.37 ± 5.83) and 55 keV (28.29 ± 16.86 vs. 9.88 ± 7.01), and M_0.6 series (11.42 ± 6.00). Pancreas-to-lesion contrast peaked in the nMERA 40 keV series (26.39 ± 16.83) and was significantly higher than in all other series (p<0.001). nMERA 55 keV images series were consistently preferred by both observers over all other series (p<0.01).

CONCLUSIONS

nMERA DECT can significantly improve image quality and pancreas-to-lesion contrast in the diagnosis of pancreatic adenocarcinoma.