Photon-Counting CT Shows Higher Spatial Resolution for Lung Nodules in Phantom Study

Advancing Precision in Pulmonary Nodule Surveillance

The evolution of computed tomography consistently focuses on enhancing image quality while minimizing radiation exposure, a balance critical for the long-term monitoring of pulmonary nodules. While conventional energy-integrating detectors have long been the clinical standard, they are limited by electronic noise and fixed spatial resolution thresholds [1, 2]. Photon-counting computed tomography (PCCT) offers an alternative detector physics model, providing intrinsic spectral data and higher spatial resolution without the penalty of electronic noise [3, 4]. These technical improvements are essential for the accurate quantification of imaging biomarkers, which are increasingly used to drive oncological treatment decisions and response assessments [5]. A new study evaluates how these competing high-resolution technologies perform in the specific task of detecting diverse lung nodules across a range of clinical dose levels, providing practical guidance for radiologists managing lung cancer screening protocols.

Comparative Phantom Modeling at Variable Dose Levels

To evaluate diagnostic performance, researchers conducted a head-to-head comparison of pulmonary nodule detectability using two distinct systems: a standard ultra-high-resolution computed tomography (UHR-CT) scanner equipped with conventional energy-integrating detectors and a photon-counting computed tomography (PCCT) scanner operated in ultra-high-resolution mode. The experimental design utilized both a standard image quality phantom and an anthropomorphic phantom (a specialized physical model designed to mimic the anatomical structures and tissue attenuation characteristics of the human torso). By using these controlled models, the team objectively measured how detector physics influence the visualization of small lung lesions without the confounding variables present in human subjects.

The study protocol involved scanning both phantoms at three specific radiation dose levels to simulate various clinical scenarios: 7.5 mGy, 2.5 mGy, and an ultra-low dose of 0.4 mGy. To ensure the findings remained applicable to daily practice, the researchers selected imaging parameters typically used for standard ultra-high-resolution chest scans. This setup allowed for the calculation of the detectability index (d'), a statistical metric that models how well a human observer can distinguish a specific nodule from background noise. By testing these systems across a range of exposures, the study determined whether the physical advantages of photon-counting technology translate into superior clinical performance at the lower radiation doses routinely required for longitudinal nodule surveillance.

Quantifying Noise Texture and Spatial Resolution

To provide an objective assessment of image quality, the researchers utilized the noise power spectrum (a mathematical tool used to quantify both the magnitude and the visual texture of image noise). Within this framework, they measured the average frequency of noise, designated as the f_av value. Additionally, the team employed a task-based transfer function to evaluate spatial resolution, specifically focusing on the f_50 value (the spatial frequency at which the imaging system maintains 50 percent contrast, serving as a critical indicator of the scanner's ability to resolve fine anatomical details). By applying these physical metrics, the study distinguished between the raw clarity of the image and the artificial smoothing often introduced by reconstruction algorithms.

The analysis revealed that spatial resolution (f_50) values were higher with PCCT than with UHR-CT across all radiation dose levels and for both phantom inserts. While both systems experienced a decrease in f_50 values as the dose was reduced from 7.5 mGy to 0.4 mGy, the photon-counting system consistently maintained superior effective spatial resolution. This performance advantage is rooted in how the two technologies handle noise. In the PCCT system, while noise magnitude increased as the dose decreased, the noise texture (f_av) remained similar, demonstrating a dose-invariant noise texture. This stability means the visual character of the image remains consistent for the radiologist reading the scan, even at lower exposures.

In contrast, the conventional UHR-CT system exhibited a different profile where both noise magnitude and noise texture (f_av) decreased as the radiation dose was lowered. This indicates that the UHR-CT system relies heavily on dose-dependent noise suppression, which results in a smoother image but at the cost of reduced spatial frequency. Because the PCCT system avoids this reliance on aggressive smoothing, it achieves a higher effective spatial resolution compared to the energy-integrating detectors of the UHR-CT. For the practicing clinician, these data suggest that PCCT provides more consistent image texture and sharper detail retention during low-dose protocols, whereas conventional ultra-high-resolution systems sacrifice fine detail to manage noise at those same levels.

Nodule Detectability and Radiologist Confidence

To bridge the gap between physical measurements and clinical utility, the researchers computed detectability indices (d') to model the detection of three specific chest nodules across the various dose levels. The results showed a notable divergence in performance between the two technologies as radiation doses changed. Specifically, detectability index (d') values increased as the dose increased with PCCT, following a standard relationship between signal and noise. Conversely, d' values decreased as the dose increased with UHR-CT, a counterintuitive finding attributed to the system's internal processing. The researchers noted that the UHR-CT relied on dose-dependent noise suppression, a software-based technique that reduces visual graininess but simultaneously produces smoother images and reduced spatial frequency. This trade-off suggests that while the UHR-CT images may appear cleaner to the eye at certain settings, the underlying mathematical detectability of fine structures is compromised by the smoothing algorithms.

Subjective assessments mirrored these objective metrics, as five radiologists evaluated their confidence in detecting nodules on the anthropomorphic phantom. For high-contrast solid nodules, the clinicians reported that confidence in nodule detection was clinically sufficient at all dose levels, confirming that high-contrast nodule detection can be performed at ultra-low dose levels with both computed tomography systems. However, the detection of more subtle lesions proved highly dose-dependent. For subsolid nodules (lesions characterized by a ground-glass appearance that partially obscures underlying lung parenchyma), radiologist confidence was clinically sufficient at 7.5 mGy and 2.5 mGy but fell below the threshold at the lowest dose level. The most significant challenge was observed with low-contrast nodules, where radiologist confidence was clinically sufficient only for PCCT at 7.5 mGy. These findings indicate that while both systems are robust for routine solid nodule screening, photon-counting technology offers a distinct clinical advantage in identifying low-contrast lesions that are often difficult to characterize on conventional energy-integrating systems, potentially allowing for earlier intervention in high-risk patients.

References

1. Tortora M, Gemini L, D’Iglio I, Ugga L, Spadarella G, Cuocolo R. Spectral Photon-Counting Computed Tomography: A Review on Technical Principles and Clinical Applications. Journal of Imaging. 2022. doi:10.3390/jimaging8040112

2. Sartoretti T, Wildberger JE, Flohr T, Alkadhi H. Photon-counting detector CT: early clinical experience review. British Journal of Radiology. 2023. doi:10.1259/bjr.20220544

3. Bie JVD, Straten MV, Booij R, et al. Photon-counting CT: Review of initial clinical results. European Journal of Radiology. 2023. doi:10.1016/j.ejrad.2023.110829

4. Greffier J, Viry A, Robert A, Khorsi M, Si‐Mohamed S. Photon-counting CT systems: A technical review of current clinical possibilities. Diagnostic and Interventional Imaging. 2024. doi:10.1016/j.diii.2024.09.002

5. Chauvie S, Mazzoni L, O’Doherty J. A Review on the Use of Imaging Biomarkers in Oncology Clinical Trials: Quality Assurance Strategies for Technical Validation. Tomography. 2023. doi:10.3390/tomography9050149