A pilot study: dual-contrast arthrography using photon-counting detector computed tomography with gadolinium and iodine contrasts

OBJECTIVES: Photon-counting detector CT (PCD-CT) enables spectral imaging and K-edge-based material separation. In arthrography, iodine and gadolinium are often co-administered for complementary CT/MR evaluation but cannot be readily separated or quantified with conventional CT or MRI. We developed a dual-contrast PCD-CT material decomposition approach to simultaneously separate and quantify iodine, gadolinium, and calcium. Approach: Calibration phantoms containing iodine, gadolinium, calcium, and iodine--gadolinium mixtures, as well as an ex vivo pig-tail specimen, were scanned on a benchtop PCD-CT system. Six energy bins spanning the iodine and gadolinium K-edges were reconstructed and analyzed using principal component analysis (PCA). Linearity, fit-based quantitative accuracy, and contrast-to-noise ratio (CNR) were evaluated for pure-agent phantoms (1--5%) and mixed-contrast phantoms (1--3%). Main results: PCA-based decomposition successfully separated iodine, gadolinium, and calcium. In phantoms containing both pure and mixed contrast, the normalized signal showed strong linearity with concentration, with slopes of 0.94 for iodine and 0.97 for gadolinium (\(R^2 = 0.989\) and 0.997, respectively). In calcium-containing mixed phantoms, linearity remained high after normalization, with slopes of 0.98 for iodine and 0.93 for gadolinium (\(R^2 = 0.99\) for both). Fit-based MAE/RMSE were 0.13%/0.16% for pure iodine and 0.04%/0.04% for pure gadolinium. Both contrast agents were detectable at 1% weight/volume and above, with CNR values of 5-8 under the current imaging conditions. Ex vivo scans showed clear separation of bone and injected contrast. Significance: This pilot study demonstrates the feasibility of PCA-based dual-contrast PCD-CT arthrography for simultaneous imaging of calcium, iodine, and gadolinium. The method provides complementary quantitative information within a dual-contrast workflow and shows promise for musculoskeletal and joint imaging.