Photon-counting detector CT with virtual monoenergetic imaging and iterative metal artifact reduction for spinal fusion implants: optimal keV and combined benefits.

Objectives

To evaluate the effectiveness of virtual monoenergetic imaging (VMI) and iterative metal artifact reduction (iMAR), alone and in combination, for reducing metal artifacts in postoperative CT scans of patients with spinal fusion, using a clinical photon-counting detector CT (PCD-CT) system.

Materials And Methods

Consecutive patients after spinal instrumentation were prospectively included in this study and received PCD-CT imaging of spine. Image reconstructions comprised conventional polychromatic images (PI), virtual monoenergetic images (VMI) at 100-140 keV, and iterative metal artifact reduction (iMAR) for determining the optimal keV. Further a combined optimal keV VMI with iMAR was reconstructed. Objective image quality was quantified by image noise and artifact index (AIx) measured across five predefined anatomical regions. Subjective assessment was performed independently by two radiologists using a 5-point Likert scale to grade artifact severity and the visibility of adjacent anatomical structures.

Results

66 patients (32 males, 34 females; 60.5 ± 14.2 years) with a total of 70 scans were enrolled. All VMI reconstruction provided lower noise, lower AIx and better visibility and diagnostic confidence for both vertebra and adjacent anatomical structures than polychromatic images (P < 0.0001). Subjective and objective assessments in first step showed that 120 keV provided superior overall imaging quality. The further combination of 120-keV VMI with iMAR resulted in significantly lower metal artifacts but introduced new artifacts.

Conclusions

For spinal implants, PCD-CT offers effective approaches for metal artifact and 120 keV is preferred. The combination of iMAR and 120 keV introduce new artifacts and requires careful consideration.