FC68 vs FC21 Reconstruction Kernels on Toshiba Aquilion
For head CT reconstruction at 0.5mm slice thickness on Toshiba Aquilion, FC21 is a smoother, standard kernel that provides lower noise and better gray-white matter contrast, while FC68 is a sharper, edge-enhancing kernel that increases spatial resolution but also increases image noise.
Key Differences
FC21 (Standard/Smooth Kernel)
- Lower spatial frequency reconstruction designed for soft tissue visualization
- Produces images with reduced noise and better contrast-to-noise ratio
- Provides superior gray-white matter differentiation 1
- Better suited for routine brain parenchymal evaluation
- More isotropic image quality when creating multiplanar reconstructions
FC68 (Sharp/Edge-Enhancing Kernel)
- Higher spatial frequency reconstruction that enhances edges and fine details
- Increases spatial resolution for detecting subtle fractures or bone detail
- Significantly increases image noise, which can obscure subtle parenchymal findings
- Creates more anisotropic image quality, particularly problematic for multiplanar reformats 2
- May reduce detectability of low-contrast lesions despite improved sharpness
Clinical Application Algorithm
For routine head CT interpretation:
- Use FC21 for primary diagnostic images - optimizes parenchymal assessment, hemorrhage detection, and gray-white differentiation 1
- Consider FC68 as a supplementary reconstruction only when:
- Evaluating skull base fractures requiring maximum bone detail
- Assessing temporal bone structures
- Creating high-resolution bone algorithm images 3
Important caveat: At 0.5mm slice thickness, both kernels maintain adequate spatial resolution for most diagnostic tasks. The choice primarily affects the noise-resolution tradeoff rather than fundamental diagnostic capability 4, 2. However, FC21's lower noise provides better overall diagnostic accuracy for brain parenchyma, while FC68's increased noise can actually decrease detectability of small or low-contrast lesions despite its sharper appearance 2.
The research demonstrates that smoother kernels like FC21 provide higher detectability indices and better task-based performance for typical diagnostic scenarios, while sharper kernels are reserved for specific high-resolution bone imaging needs 1, 2.