Voxel-wise deep learning segmentation of hydroxyapatite and iodine in spectral photon-counting CT: A quantitative phantom study

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Dice coefficient 0.72 ± 0.01 for SPFF-UNet vs 0.66 ± 0.02 for ResUNet++

• Primary Outcome: Thirteen-class voxel-wise segmentation performance

• Nnt Or Sensitivity: Sensitivity 0.73 ± 0.01, Precision 0.71 ± 0.01 for SPFF-UNet

• Confidence Interval: Standard deviations reported (±0.01-0.03) but confidence intervals not specified

Clinical Application

Requires spectral photon-counting CT technology and computational resources for deep learning implementation Relevant for Australian radiology departments with advanced CT capabilities; potential applications in cardiovascular risk assessment and musculoskeletal imaging aligned with RACGP preventive care guidelines Patients requiring differentiation of hydroxyapatite deposits from iodinated contrast in calcific musculoskeletal disease and vascular calcification

Abstract

Accurate non-invasive identification of hydroxyapatite (HA) deposits is important for diagnosing calcific musculoskeletal disease and quantifying vascular calcification, but conventional and dual-energy CT often struggle to distinguish HA from iodinated contrast because of overlapping attenuation, noise, and beam-hardening artifacts. Spectral photon-counting CT (SPCCT) offers improved energy resolution and spatial fidelity, yet most deep-learning approaches in spectral CT focus on continuous density regression or anatomical segmentation rather than direct voxel-wise material labeling. We developed SPFF-UNet, a spectral-preserving 3D segmentation model for direct classification of HA and iodine concentrations from five-bin SPCCT volumes without material-decomposition preprocessing. A cylindrical phantom containing twelve materials was scanned at 0.1 mm isotropic resolution, including five HA concentrations, three iodine concentrations, three soft-tissue equivalents, and water. SPFF-UNet integrates spectral squeeze-excitation, EnergyFiLM, and FourierGate to preserve and exploit multi-energy information throughout the network. The model was trained for thirteen-class voxel-wise segmentation and compared with five established 3D architectures under matched training conditions. SPFF-UNet achieved the best macro-averaged performance on a held-out phantom scan (Dice 0.72 ± 0.01, IoU 0.59 ± 0.01, sensitivity 0.73 ± 0.01, precision 0.71 ± 0.01), outperforming the strongest comparator, ResUNet++ (Dice 0.66 ± 0.02, IoU 0.46 ± 0.02, sensitivity 0.67 ± 0.02, precision 0.61 ± 0.03). Performance gains were concentrated in mid/low-contrast HA and low-concentration iodine, with reduced slice-wise variability and fewer HA-iodine misclassifications. These results suggest that preserving spectral information and applying targeted spectral modulation can improve concentration-aware voxel classification from SPCCT. This phantom-based proof-of-concept provides a basis for future in vivo validation.

References

1. 1.Francis, N., Seghier, M. L., Maalej, N., & Raja, A. (2026). Voxel-wise deep learning segmentation of hydroxyapatite and iodine in spectral photon-counting CT: A quantitative phantom study. PLoS One. https://doi.org/10.1002/mp.16363