Spinal dual-energy computed tomography: improved visualisation of spinal tumorous growth with a noise-optimised advanced monoenergetic post-processing algorithm

Mareen Kraus; Jakob Weiss; Nadja Selo; Thomas Flohr; Mike Notohamiprodjo; Fabian Bamberg; Konstantin Nikolaou; Ahmed E Othman

doi:10.1007/s00234-016-1733-7

Spinal dual-energy computed tomography: improved visualisation of spinal tumorous growth with a noise-optimised advanced monoenergetic post-processing algorithm

Neuroradiology. 2016 Nov;58(11):1093-1102. doi: 10.1007/s00234-016-1733-7. Epub 2016 Aug 11.

Authors

Mareen Kraus¹, Jakob Weiss¹, Nadja Selo¹, Thomas Flohr², Mike Notohamiprodjo¹, Fabian Bamberg¹, Konstantin Nikolaou¹, Ahmed E Othman³

Affiliations

¹ Department of Diagnostic and Interventional Radiology, Eberhard Karls University Tuebingen, Hoppe-Seyler-Straße 3, 72076, Tuebingen, Germany.
² Siemens Healthcare GmbH, Erlangen, Germany.
³ Department of Diagnostic and Interventional Radiology, Eberhard Karls University Tuebingen, Hoppe-Seyler-Straße 3, 72076, Tuebingen, Germany. Ahmed.e.Othman@googlemail.com.

PMID: 27516101
DOI: 10.1007/s00234-016-1733-7

Abstract

Introduction: The aim of this study was to evaluate the effect of advanced monoenergetic post-processing (MEI+) on the visualisation of spinal growth in contrast-enhanced dual-energy CT (DE-CT).

Methods: Twenty-six oncologic patients (age, 61 ± 17 years) with spinal tumorous growth were included. Patients underwent contrast-enhanced dual-energy CT on a third-generation dual-source CT scanner. Image acquisition was in dual-energy mode (100/Sn150kV), and scans were initiated 90 s after contrast agent administration. Virtual monoenergertic images (MEI+) were reconstructed at four different kiloelectron volts (keV) levels (40, 60, 80, 100) and compared to the standard blended portal venous computed tomography (CT_pv). Image quality was assessed qualitatively (conspicuity, delineation, sharpness, noise, confidence; two independent readers; 5-point Likert scale; 5 = excellent) and quantitatively by calculating signal-to-noise (SNR) and contrast-to-noise-ratios (CNR). For a subgroup of 10 patients with MR imaging within 4 months of the DE-CT, we compared the monoenergetic images to the MRIs qualitatively.

Results: Highest contrast of spinal growth was observed in MEI+ at 40 keV, with significant differences to CT_pv and all other keV reconstructions (60, 80, 100; p < 0.01). Highest conspicuity, delineation and sharpness were observed in MEI+ at 40 keV, with significant differences to CT_pv (p < 0.001). Similarly, MEI+ at 40 keV yielded highest diagnostic confidence (4.6 ± 0.6), also with significant differences to CT_pv (3.45 ± 0.9, p < 0.001) and to high keV reconstructions (80, 100; p ≤ 0.001). Similarly, CNR calculations revealed highest scores for MEI+ at 40 keV followed by 60 keV and CT_pv, with significant differences to high keV MEI+ reconstructions. Qualitative analysis scores peaked for MR images followed by the MEI+ 40-keV reconstructions.

Conclusion: MEI+ at low keV levels can significantly improve image quality and delineation of spinal growth in patients with portal-venous phase CT scans due to increased CNR and limited image noise.

Keywords: Dual-energy CT; Monoenergetic post-processing; Spinal tumour.

Publication types

Evaluation Study

MeSH terms

Algorithms*
Artifacts
Female
Humans
Male
Middle Aged
Radiographic Image Enhancement / methods*
Radiography, Dual-Energy Scanned Projection / methods*
Reproducibility of Results
Sensitivity and Specificity
Signal-To-Noise Ratio
Spinal Cord Neoplasms / diagnostic imaging*
Spinal Cord Neoplasms / pathology
Spinal Neoplasms / diagnostic imaging*
Spinal Neoplasms / pathology