Optimization of the reconstruction kernel for temporal bone imaging using photon-counting detector CT: A combined...

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Photon-counting detector CT (PCD-CT) offers enhanced spatial resolution and noise performance over conventional CT.
• Optimal reconstruction parameters, particularly kernel selection, for temporal bone imaging with PCD-CT are not well-established.

Purpose of the Study:

• To determine the optimal reconstruction kernel for temporal bone imaging using PCD-CT.
• To evaluate kernels using objective image quality metrics and subjective expert assessments.

Main Methods:

• System performance function (SPF) analysis, including task-based transfer function (TTF) and noise power spectrum (NPS), was performed on 11 kernels.
• Radiologists visually assessed image quality and visualization of fine anatomical structures (stapes footplate, incudomalleolar joint, cochlea) using a subset of kernels.

Main Results:

• TTF analysis showed Hr96 offered the highest spatial resolution, while NPS indicated Hr60 had the lowest noise.
• SPF analysis identified Hr72 as the optimal balance between resolution and noise.
• Expert visual assessment favored Hr76 for overall image quality and fine structure visualization, with Hr60 rated lowest (p < 0.05).

Conclusions:

• The Hr76 reconstruction kernel is suitable for middle and inner ear diagnoses using PCD-CT.
• Hr76 provides an effective balance between spatial resolution and image noise for high-resolution temporal bone imaging.
• These findings support the standardization of reconstruction protocols for clinical PCD-CT practice.