Reducing artifacts from total hip replacements in dual layer detector CT: Combination of virtual monoenergetic images and orthopedic metal artifact reduction. Issue 111 (February 2019)

• Record Type: Journal Article
• Authors: Neuhaus, Victor; Grosse Hokamp, Nils; Zopfs, David; Laukamp, Kai; Lennartz, Simon; Abdullayev, Nuran; Maintz, David; Borggrefe, Jan
• Is Part Of: European journal of radiology. Issue 111(2019)
• Page Start: 14
• Page End: 20
• Publication Date: 2019-02
• Subjects: Dual-energy CT -- Dual-layer CT -- Metal artifact reduction -- Virtual monoenergetic images -- Iterative metal artifact reduction; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2018.12.008
• Languages: English
• ISSNs: 0720-048X

Abstract

Abstract: Purpose: To evaluate the reduction of artifacts caused by total hip replacements (THR) in dual-layer DECT (DLCT) provided by the combination of virtual monoenergetic images (VMI) and orthopedic metal artifact reduction (MAR). Materials and Methods: A total of 24 consecutive patients carrying THR, who received DLCT, were included. Four different images were reconstructed from the same CT dataset: a) conventional images (CI), b) conventional images with orthopedic metal artifact reduction (CIMAR ) c) VMI and d) VMI combined with orthopedic metal artifact reduction (VMIMAR ). VMI and VMIMAR were reconstructed at 140 keV, 160 keV, 180 keV and 200 keV. Attenuation (HU) and noise (SD) were measured in order to evaluate reduction of hypodense and hyperdense artifacts, evaluate reduction of image noise as well as to calculate contrast-to-noise ratios (CNR). Image quality was additionally rated with regard to: a) extent of artifact reduction and assessment of b) pelvic organs, c) bone and d) muscle adjacent to the metal implants. Statistical analysis was performed using Wilcoxon test. Results: VMIMAR at high keV, 140, 160, 180 and 200 keV, led to the greatest reduction of hypodense artifacts in comparison to plain VMI or CIMAR (p < 0.01), while in comparison to CI hyperdense artifacts were significantly reduced in all reconstructions (p < 0.05). Accordingly, subjective analysis found VMIMAR to be superior in reducing hypodense artifacts in comparison to VMI and CIMAR (p < 0.05), while hyperdense artifacts were equally reduced in all reconstructions compared to CI (p < 0.0001). Additionally, assessment of the pelvic organs and adjacent bone was significantly improved in VMIMAR in comparison to VMI and CIMAR (p < 0.05). In contrast, muscles adjacent to the metal implants were significantly better assessable in all reconstructions compared to CI (p < 0.01). Conclusion: The combination of VMI and MAR yields strongest reduction of hypo- and hyperdense artifacts caused by total hip replacements in staging DLCT in comparison to each technique by itself.