Performance of dual layer dual energy CT virtual monoenergetic images to identify early ischemic changes in patients with anterior circulation large vessel occlusion

Highlights

• Virtual monoenergetic images impact identification of early ischemic changes on CT.

• Automated 70 keV ASPECTS have the highest diagnostic accuracy and sensitivity.

• Automated 40–50 keV ASPECTS show low sensitivity, especially in cortical regions.

Abstract

Background and purpose

Dual energy CT is increasingly available and used in the standard diagnostic setting of ischemic stroke patients. We aimed to evaluate how different dual energy CT virtual monoenergetic energy levels impact identification of early ischemic changes, compared to conventional polyenergetic CT images.

Materials and methods

This retrospective single-center study included patients presenting with acute ischemic stroke caused by an occlusion of the intracranial internal carotid artery or proximal middle cerebral artery. Data was gathered on consecutive patients admitted to our institution who underwent initial diagnostic stroke imaging with dual layer dual energy CT and a subsequent follow-up CT one to three days after admission. Automated ASPECTS results from conventional polyenergetic and different virtual monoenergetic energy level reconstructions at admission were generated and compared to reference standard ASPECTS. Confidence intervals (CI) for sensitivity, specificity, negative and positive predictive value were calculated.

Results

A total of 24 patients were included. Virtual monoenergetic reconstructions of 70 keV had the highest region-based ASPECTS accuracy, 0.90 (sensitivity 0.82 (95% CI 0.72–0.93), specificity 0.92 (0.88–0.97), negative predictive value 0.94 (0.90–0.96)), whereas virtual monoenergetic reconstructions of 40 keV had the lowest, 0.77 (sensitivity 0.34 (0.26–0.42), specificity 0.90 (0.89–0.96), negative predictive value 0.80 (0.77–0.83)).

Conclusions

Automated 70 keV ASPECTS had the highest diagnostic accuracy, sensitivity and negative predictive value overall. Our results indicate that virtual monoenergetic energy levels impact the identification of early ischemic changes on CT.