(2026) Bassermann | PCCT reduces beam-hardening effects in proton stopping-power estimation

Ivo Driesser

This phantom study compared photon-counting CT (PCCT), dual-energy CT (DECT), and single-energy CT (SECT) for proton stopping-power ratio (SPR) estimation, specifically testing how phantom size and insert position affect accuracy through beam hardening. Two SPR methods were evaluated: a DECT-based Saito method and a conventional Hounsfield look-up table (HLUT) method.

The principal finding was that PCCT combined with VMI-based HLUT estimation was the most stable approach, particularly in larger phantoms. For PCCT, non-bone SPR deviations with HLUT were essentially within ±1%, while peripheral errors became substantially larger for DECT and SECT.

In the most challenging 40-cm phantom, PCCT + HLUT limited peripheral deviation to +2.1% for cortical bone, whereas deviations reached −5.5% for DECT/SECT configurations. Overall, the authors conclude that PCCT-VMI + HLUT is less susceptible to size- and position-dependent beam hardening than DECT or SECT.

Access publication

Protocol

Detail	Value
PCCT scanner	NAEOTOM Alpha
DECT / SECT scanner	SOMATOM Definition Edge
Phantoms	Gammex Advanced Electron Density (AED) phantom + custom Modular Evaluation (ME) phantom
ME phantom diameters	10, 20, 30, 40 cm
Tissue-equivalent inserts	12 cylindrical inserts, 2.8 cm diameter × 16.5 cm length
Beam-hardening test inserts	CaCO₃ 50% and HE Cortical Bone
PCCT tube voltage	140 kVp
DECT acquisition	Dual-spiral, 80/140 kVp
SECT tube voltage	120 kVp
PCCT reconstruction	QIR strength 3
DECT / SECT reconstruction	ADMIRE strength 3
Kernel	Qr40 for all modalities
PCCT beam-hardening correction	iBHC Quantum
DECT / SECT beam-hardening correction	iBHC bone
Matrix	512 × 512
Slice thickness	2 mm
Voxel size range	0.29 × 0.29 × 2.00 to 0.88 × 0.88 × 2.00 mm³
PCCT protocol	Standard abdominal protocol
PCCT ME phantom CTDIvol	10.2 mGy
DECT ME phantom CTDIvol	20.1 mGy low dose; 35.0 mGy high dose
SECT ME phantom CTDIvol	19.7 mGy low dose; 35.1 mGy high dose
VMI range	40–190 keV, 10-keV increments
Number of VMI levels	16
VOI	1.9 cm diameter × 2.0 cm length
SPR methods	Saito/Sagara DECT-based method + SECT-based HLUT
PCCT HLUT VMI selected	150 keV small phantom; 170 keV large phantom
PCCT Saito VMI pair selected	150/160 keV small phantom; 80/90 keV large phantom
Primary evaluation	Size- and position-dependent SPR deviation / beam hardening

The acquisition and reconstruction parameters are reported directly in the study's image-acquisition section. The AED phantom provided calibration data, whereas the multi-diameter ME phantom was used for evaluation.

Key Results

Parameter	Main finding
Overall best combination	PCCT VMI + HLUT
PCCT HLUT calibration RMSD	0.5% at both selected VMI energies: 150 keV (small) and 170 keV (large)
Non-bone inserts, PCCT + HLUT	Within ±1.0% across all ME diameters except HE Liver at 10 cm (−1.1%)
CaCO₃ 50%, size dependence	PCCT + HLUT deviations <0.5%
HE Cortical Bone, size dependence	PCCT + HLUT deviation up to 0.9%, lowest among compared modalities
20-cm phantom, position dependence	All modalities/methods remained within 2.3%
30-cm phantom, HLUT	PCCT, high-dose DECT and SECT remained within 1.7%
40-cm phantom	Strongest position-dependent beam-hardening effects
40-cm peripheral cortical bone, PCCT + HLUT	Maximum deviation +2.1%
Largest peripheral deviations	Up to −5.5% for high-dose SECT and low-dose DECT with the Saito method
HLUT vs Saito	HLUT generally produced smaller and more stable SPR deviations
Dose observation	PCCT ME scans used 10.2 mGy, versus ~20 mGy for low-dose DECT/SECT and ~35 mGy for high-dose acquisitions
Main implication	PCCT-VMI + HLUT reduced the influence of beam hardening and produced more stable proton SPR estimates across object sizes and positions.

An important nuance is that the study did not find a clear correlation between phantom diameter and estimated SPR across all configurations; the major advantage of PCCT emerged especially in position-dependent stability in larger objects. The authors also caution that this was a simplified phantom experiment involving only one scanner of each type, so the findings cannot automatically be generalized to all clinical scanners or patient geometries.

Source:
Bassermann L, Edmund JM, Olech WE, et al. Impact of beam hardening on proton stopping-power ratio estimation using photon-counting, dual-energy, and single-energy computed tomography. Physics and Imaging in Radiation Oncology. Published online August 17, 2026; article 101064. doi: 10.1016/j.phro.2026.101064.