Improved assessment of the cochlear implant and temporal bone structures with ultra-high resolution photon-counting detector CT compared to conventional CT systems

PURPOSE: This study evaluates the image quality of photon-counting detector CT (PCD-CT) compared to conventional CT systems in cochlear implant imaging and aims to determine optimal scan parameters. METHODS: Three cochlear implant models were inserted into three post-mortem human specimen heads. Each head was scanned on three high-end CT systems: SOMATOM Edge, SOMATOM Force and NAEOTOM Alpha. A total of 192 reconstructions were obtained, including various tube voltages (120-150 kV), tin filtering, and slice thicknesses (0.2-0.6 mm). For each scanner, multiple reconstruction kernels were applied. Contrast-to-noise ratios (CNR) for bone-to-soft-tissue were calculated. Two radiologists independently assessed anatomy, implant visibility, and overall image quality on a 5-point Likert scale and scored diagnostic confidence for detecting anatomic dehiscence. RESULTS: PCD-CT showed significantly higher image quality scores across implant, anatomy, and general image quality compared to the Edge and Force (p < 0.001). PCD-CT also produced significantly higher CNR (mean 23.5 vs. 13.8 [Edge] and 12.7 [Force]) and greater diagnostic confidence. For PCD-CT, optimal image quality was achieved using 0.2 mm slice thickness and the Hr89 kernel at 140 kV. PCD-CT scans acquired with half dose (27.5 mGy) showed no significant difference in CNRs but led to slightly lower subjective scores compared to standard-dose imaging. CONCLUSION: PCD-CT provides superior image quality and diagnostic confidence in cochlear implant imaging compared to the Edge and Force. Optimal results are achieved using 140 kV, 0.2 mm slices, and the Hr89 kernel. Although dose reduction is feasible, it slightly compromises subjective image quality.