RESUMO
Objective:To access the efficacy of monoenergetic imaging from spectral CT combined with metal artifact reduction for orthopedic implants (O-MAR) on reducing contrast hardening artifacts in the vein on the injection side, and determining the optimal monoenergetic spectral range to improve the display of axillary lymph node.Methods:A total of 35 patients with breast cancer who underwent chest-enhanced CT scans were enrolled in this retrospective study. The original data were reconstructed to obtain a total of 35 sets of images, including one conventional image, 17 groups of monoenergetic images, and 17 groups of monoenergetic+O-MAR images. The areas of interest were delineated in the high and low-density artifact area on the injection side of the same layer contrast agent, and the contralateral ectopectoralis. The CT value and its standard deviation (SD) were recorded respectively, the artifact area was measured, and the number of axillary lymph nodes was recorded. The difference in CT values (ΔCT 1, ΔCT 2) and the artifact index (AI1 and AI 2) of the high and low-density artifact areas relative to the contralateral ectopectoralis in the same layer were calculated respectively. Friedman test and Wilcoxon signed-rank test were used to compare the differences of ΔCT, AI, artifact area, and number of lymph nodes among the three imaging modalities, and the Kappa test was used to compare the differences in subjective evaluation. Results:As the energy level increased, compared to the conventional image, monoenergetic image, ΔCT 1 absolute value, ΔCT 2 absolute value, AI 1, and AI 2 showed a trend of initially low and then high, artifact area decreased, and the number of detected lymph nodes increased ( P<0.01). Compared to other energy levels, when the monoenergetic image was 100 keV, ΔCT 1 value, 140 keV for ΔCT 2 value, 120 keV for AI 1 value, and 130 keV for AI 2 value were close to zero, and the number of detected lymph nodes was highest at 110-200 keV. In contrast, in the monoenergetic+O-MAR images, ΔCT 1 absolute value showed a trend of initially low and then high, but, ΔCT 2 absolute value, AI 1, AI 2, and artifact area all significantly decreased, whereas the number of detected lymph nodes significantly increased (χ 2 values were 916.23, 895.93, 387.08, 519.41, 890.10, and 1027.98, respectively. All P<0.01). Compared to other energy levels, when the monoenergetic+O-MAR image was at 100 keV, ΔCT 1 value was close to zero, while ΔCT 2 value became close to zero with increasing energy level, and the number of detected lymph nodes was highest at 110-200 keV. As the energy level increased, the ΔCT 1, AI 1, AI 2, and artifact area of monoenergetic+O-MAR images were significantly smaller than those of monoenergetic images at the same energy level, and the number of detected lymph nodes was significantly higher than that of monoenergetic images ( P<0.01). The subjective scores for 110-200 keV monoenergetic images and 100-200 keV monoenergetic+O-MAR images were both higher than 4, and the score for monoenergetic+O-MAR images was significantly higher than that of single-energy spectrum images. The agreement between the two radiologists in assessing subjective scores was good. Conclusion:At 100-120 keV level, spectral CT monoenergetic combined with O-MAR imaging technique has the best performance in removing hardening-induced artifacts of chest-enhanced CT contrast agent and detecting and displaying axillary lymph nodes.
