Note: This abdominal imaging deep dive paper does not specify scan protocols, but give an overview on clinical practice.
Access publication
Figure 2. Early-stage pancreatic cancer was evaluated using a photon-counting detector CT. A 57-year-old female patient with suspected early-stage pancreatic cancer based on focal pancreatic atrophy on magnetic resonance cholangiopancreatography underwent dynamic contrast-enhanced PCD-CT. In the arterial phase, 70 keV VMI (a) demonstrates focal atrophy and a subtle hypoattenuating area in the pancreatic tail. The lesion remains inconspicuous on 40 keV VMI (b), making its identification as a hypovascular tumor confidently challenging. In the equilibrium phase, faint delayed improvement is suspected on 70 keV VMI (c), which becomes more conspicuous on 40 keV VMI (d) (yellow arrows). Cinematic volume-rendered images reconstructed from arterial-phase 40 keV VMI data (e, f) clearly depict focal atrophy of the pancreatic tail (arrowheads). Detecting early-stage pancreatic cancer may be challenging in the arterial phase; therefore, findings, including focal parenchymal atrophy and delayed enhancement, are important. PCD-CT is particularly advantageous for detecting these subtle features
Partially AI-generated overview
UHR PCCT in abdominal imaging
- UHR PCCT enables slice thicknesses down to 0.2 mm, improving visualization of small vessels, pancreatic ducts, urinary calculi, and other fine anatomical structures while reducing blooming artifacts.
- The greatest clinical benefit is seen in CT angiography, interventional planning, urinary stone detection, and evaluation of calcified or high-contrast structures, whereas improvements for low-contrast abdominal lesions remain more limited.
- A dual-reconstruction workflow is recommended, using standard-resolution images for routine interpretation and targeted UHR reconstructions for problem-solving applications requiring maximum spatial detail (see reconstruction).
Spectral imaging
- Spectral imaging is inherently available in every PCCT examination with specific scan modes, enabling retrospective reconstruction of virtual monoenergetic images (VMIs), iodine maps, and material decomposition without dedicated dual-energy acquisition.
- Low-keV VMIs (40–50 keV) substantially enhance iodine contrast, improving visualization of hypervascular and hypovascular lesions, while 70 keV closely resembles conventional CT and serves as the preferred routine reconstruction.
- Spectral imaging also supports virtual non-contrast (VNC) imaging, quantitative iodine mapping, and contrast-media reduction strategies while maintaining diagnostic image quality.
Clinical impact of spectral imaging
- Low-keV VMIs improve detection and characterization of pancreatic cancer, liver metastases, rectal cancer, inflammatory diseases, and vascular abnormalities by increasing lesion conspicuity and tissue contrast.
- Quantitative iodine density maps provide functional information that may serve as imaging biomarkers for tumor perfusion, treatment response, inflammatory activity, and lymph node assessment.
- The combination of high-resolution anatomical imaging and routine spectral information supports the concept of "one-stop imaging", potentially reducing additional CT or MRI examinations in selected clinical scenarios.
Reconstruction strategies and optimization
- 70-keV VMI with medium-soft kernels is recommended for routine abdominal interpretation because it closely matches conventional CT while preserving access to spectral information.
- Task-specific reconstructions using 40–50 keV VMIs, sharper kernels, thinner slices, and higher Quantum Iterative Reconstruction (QIR) levels improve lesion conspicuity and vascular visualization when clinically indicated.
- Optimizing kernel selection, slice thickness, matrix size, and QIR is essential to balance spatial resolution, image noise, and diagnostic confidence for different abdominal imaging tasks.
"Accordingly, dual reconstruction is common in clinical practice. A routine reconstruction consisting of 70-keV VMI, a medium-smooth vascular kernel (Br40–44 or Bv36), moderate QIR (QIR-2 or 3), and a slice thickness of ~0.6 mm provide stable texture and noise control for routine interpretation. Additional task-specific reconstructions may employ lower-keV VMIs (e.g., 40 keV), sharper kernels (Bv72), higher QIR (QIR-4), thinner slices (~0.2 mm), and larger matrices (e.g., 1024, 2048) for problem-solving or high-resolution evaluation."
Implementation considerations: data management and infrastructure
- Routine PCCT generates substantially larger datasets than conventional CT due to thin-slice acquisitions, high-resolution matrices, and multiple spectral reconstructions, requiring significantly greater storage and computational resources.
- Successful implementation requires optimized PACS infrastructure, selective archiving strategies, network capacity, and efficient spectral data management to prevent workflow bottlenecks.
- Institutions should view PCCT implementation as a system-wide workflow transformation, rather than simply replacing an existing CT scanner.
Future perspectives
- Future multi-energy PCCT systems are expected to expand quantitative imaging through improved material decomposition, electron density mapping, extracellular volume (ECV) assessment, and K-edge imaging.
- Emerging applications include molecular imaging, virtual non-calcium imaging, automated tissue characterization, and AI-assisted quantitative biomarkers, extending CT beyond morphology-based diagnosis.
- Wider adoption will depend on demonstrating clinical value, cost-effectiveness, multivendor standardization, and robust quantitative reproducibility across different PCCT platforms.
Source: Yokoyama K, Kawasaki Y, Fujii M, et al. Abdominal Radiology. 2026.
