Dual energy CT of peripheral arteries: effect of automatic bone and plaque removal on image quality and grading of stenoses

Eur J Radiol. 2008 Dec;68(3):414-22. doi: 10.1016/j.ejrad.2008.09.016. Epub 2008 Oct 28.

Abstract

Purpose: To evaluate the effect of automatic bone and plaque removal on image quality and grading of steno-occlusive lesions in patients undergoing dual energy CT angiography (CTA) of lower extremity.

Materials and methods: Dual energy (DE) runoff CTA was performed in 50 patients using the following parameters: collimation 2 x 32 x 0.6; tube potentials, 80 kV and 140 kV; reconstructed slice thickness 1mm. 100 mL iomeprol 400 and 50 mL saline were injected at 4 mL/s. Separate datasets were calculated for each of the two tubes and used to generate automatically bone-subtracted images (ABS) as well as bone and plaque subtracted images (ABPS). Residual bone in the ABS dataset was removed manually (=ABS-B dataset). In addition, a weighted average dataset from both dual energy acquisitions resembling a routine 120 kV CT acquisition was used for standard manual bone subtraction (MBS). Operator time for bone removal was measured. Effectiveness of bone subtraction and presence of vessel erosions was assessed by two readers in consensus. Stenosis grading in plaque subtracted and unsubtracted images was assessed and correlated.

Results: Residual bone fragments (ribs: 46%, patella: 25%, spine: 4%, pelvis: 2%, tibia 2% of patients) were only observed with ABS. The time needed to manually remove these residual bones was 2.1+/-1.1 min and was significantly lower than the duration of manual bone removal (6.8+/-2.0 min, p<0.0001, paired t-test). A total of 1159 arteries were analyzed. Compromising vessel erosions were observed less frequently in the ABS-B dataset (10.6%) than in the MBS dataset (15.2%, p<0.001, wilcoxon's signed rank test). A total of 817 steno-occlusive lesions were assessed. While the agreement of grading of steno-occlusive lesions was good at the levels of the aorta and the pelvic arteries (kappa=0.70 in both, Cohen's kappa statistics), it was moderate at the level of the thigh arteries (kappa=0.57) and poor at the level of the calf (kappa=0.16).

Conclusion: DE CTA has substantial advantages over conventional CTA. Automatic bone subtraction is more time efficient and reliable. Automatic plaque subtraction for the first time provides a true CTA-luminogram which is easy to interpret and reduces the need for further post-processing. DE CTA provides best results in arteries of the thigh; below the knee, plaque subtraction is less accurate.

Publication types

  • Evaluation Study

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Algorithms
  • Angiography / methods*
  • Arterial Occlusive Diseases / diagnostic imaging*
  • Artificial Intelligence
  • Bone and Bones / diagnostic imaging*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Pattern Recognition, Automated / methods*
  • Peripheral Vascular Diseases / diagnostic imaging*
  • Radiographic Image Enhancement / methods*
  • Radiographic Image Interpretation, Computer-Assisted / methods
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Subtraction Technique*