gms | German Medical Science

Objective: We previously reported on the beneficial value of 2^nd generation dual energy CT-angiography (DE-CTA) imaging as a non-invasive and safe imaging alternative for patients undergoing surgical or endovascular intracranial aneurysm (IA) repair. Dual-energy CTA (DE-CTA) imaging is derived from simultaneous use of two x-ray tubes, which run at different energies. In contrast to single-energy CTA (SE-CTA), this allows for calculation and image reconstruction at different tube voltages (monoenergetic images). Thus, beam hardening artifact reduction (high tube voltages) while maintaining diagnostic vessel contrast (low tube voltages) may be achieved. We here investigate the potential of virtual monoenergetic images, derived from the most recent 3^rd generation DE-CTA scanner to improve evaluation of the degree of IA occlusion and parent vessel patency.

Methods: 15 patients prospectively underwent DE-CTA imaging following surgical IA repair. Virtual monoenergetic datasets from 40-190keV and mixed 80/150keV images were calculated and compared to the respective 80keV dataset, comprising the standard SE-CTA 120kV tube voltage setting, as the reference standard. Signal to noise (SNR), contrast to noise (CNR), and signal to background (SBR) ratios were calculated in 10keV intervals ranging from 40-190keV as objective markers of image quality. Diagnostic image quality was independently rated by a radiologist and a vascular neurosurgeon with respect to IA neck and parent vessel using a 4-point likert scale from 4 (excellent) to 1 (poor). A score of 2 or higher was considered diagnostic.

Results: 15 patients surgically treated for 5 ruptured and 10 unruptured IA received DE-CTA imaging. Aneurysm sites were anterior (n=4) or posterior communicating (n=4) and middle cerebral artery (n=7), with a mean aneurysm size of 8.4 mm. All calculated ratios significantly decreased with increasing keV levels from 23.0 to 11.4 (SNR), 22.1 to 10.4 (CNR), and 26.8 to 11.4 (SBR) at 40keV to 190keV. Compared to the standard SE-CTA setting (80keV), evaluation of parent vessel patency was rated superior based on the virtual monoenergetic datasets (4 [range 2-4] vs. 3 [range 1-4]) and assessment of IA occlusion was rated superior based on the mixed 80/150keV dataset (4 [range 2-4] vs. 3 [range 1-4]). Based on the virtual monoenergetic datasets image quality was rated diagnostic in all patients, whereas image quality was rated not diagnostic in 3 patients on the standard SE-CTA (80keV) dataset.

Conclusion: Our data highlight that calculation of virtual monoenergetic datasets from DE-CTA, without additional radiation exposure, allows for case tailored tuning of the best possible vessel contrast to beam hardening artifact ratio. This may result in diagnostic imaging in patients where otherwise catheter angiography would be necessary for further evaluation.