Related Experiment Video
Updated: Aug 7, 2026
08:55
Cryo-Electron Tomography Remote Data Collection and Subtomogram Averaging
Published on: July 12, 2022
Small-pixel Acquisition Unmasks "Bucket Effects" of Iterative Reconstruction in Ultra-High-Resolution Photon-Counting
Henner Huflage1, Tobias Wech1, Theresa Sophie Patzer1
1Department of Diagnostic and Interventional Radiology, University Hospital Würzburg, Oberdürrbacher Straße 6, 97080 Würzburg, Germany (H.H., T.S.P., C.B., P.P., P.G., A.W., A.S.K., T.A.B., J-P.G.).
Academic Radiology
|August 5, 2026
Summary
Ultra-high-resolution (UHR) photon-counting CT (PCCT) imaging offers improved signal-to-noise ratio (SNR) with smaller detector pixels, especially with sharper reconstruction kernels. This allows for enhanced image quality in midface scans without increased radiation dose.
Area of Science:
- Medical Imaging
- Radiology
- Computed Tomography
Background:
- Photon-counting CT (PCCT) enables ultra-high-resolution (UHR) imaging without increased radiation dose.
- The interplay between reconstruction kernels and iterative reconstruction (IR) in UHR PCCT is not fully understood.
- Smaller detector pixels in UHR CT improve spatial sampling but their effect on SNR and image quality needs further investigation.
Purpose of the Study:
- To investigate the influence of smaller detector pixels on signal-to-noise ratio (SNR) and subjective image quality in midface PCCT.
- To evaluate the interaction between different reconstruction kernels and iterative reconstruction (IR) in UHR PCCT.
- To determine if UHR PCCT can achieve better image quality at lower radiation doses.
Main Methods:
- Cadaveric heads were scanned using UHR (120 × 0.2 mm) and standard collimation (144 × 0.4 mm) PCCT across various dose levels.
- Data were reconstructed using dedicated head kernels (Hr64, Hr76, Hr84) with varying spatial frequencies.
- Signal-to-noise ratios (SNR) were calculated, and subjective image quality was assessed by nine radiologists via pairwise comparisons.
Main Results:
- UHR datasets reconstructed with the sharper Hr76 kernel showed significantly higher SNR than standard datasets at all dose levels (p<0.001).
- Increasing kernel sharpness in UHR acquisitions did not result in the expected SNR penalty; Hr76 showed comparable or higher SNR than Hr64.
- Radiologists consistently preferred sharper UHR reconstructions, demonstrating substantial inter-reader agreement (Fleiss κ=0.711).
Conclusions:
- Smaller detector pixels in midface PCCT provide a measurable SNR advantage when matched with appropriate reconstruction sharpness.
- Kernel-dependent interactions with IR can lead to non-monotonic noise behavior, enabling sharper reconstructions without a noise penalty.
- UHR PCCT offers potential for improved image quality and diagnostic confidence in midface imaging.
