Photon-counting CT (PCCT) is the biggest change in CT detector design in decades, and it is now appearing in hospitals, journals and physics exams. This is the plain-language version.
What is different about the detector?
| Conventional CT (energy-integrating) | Photon-counting CT | |
|---|---|---|
| Conversion | Indirect: X-rays hit a scintillator, which makes light, which a photodiode turns into charge | Direct: a semiconductor (cadmium telluride, CZT or deep silicon) turns each photon straight into charge |
| What is recorded | The total energy of all photons in each reading | Each photon is counted and sorted by energy into bins |
| Electronic noise | Adds to the signal, worst at low dose | Rejected by an energy threshold |
| Spectral data | Needs dual-source, kV switching or dual-layer designs | Available from every scan |
| Detector elements | Larger, with reflective septa between them | Smaller, with no septa needed |
The first whole-body photon-counting CT was cleared by the US FDA in 2021. In 2026 GE HealthCare announced FDA clearance for a deep-silicon system with eight energy bins, so more vendors and designs are arriving.
Why it matters clinically
- Higher spatial resolution: smaller detector pixels show fine structures such as temporal bone, lung interstitium, small vessels and stents better.
- Lower noise at low dose: because electronic noise is rejected, low-dose and paediatric scans benefit most.
- Spectral imaging every time: virtual monoenergetic images, iodine maps, virtual non-contrast images and material decomposition without choosing a special protocol.
- Less blooming: calcified plaque and stents look smaller and sharper, which helps coronary CT.
- Potential contrast savings: low-keV images boost iodine signal, which may allow less contrast.
- K-edge imaging: energy bins can separate materials such as iodine and gadolinium, and new contrast agents are being researched.
Limitations
| Problem | What it means |
|---|---|
| Pulse pile-up | At very high count rates two photons arrive too close together and are counted as one higher-energy photon |
| Charge sharing | One photon's charge spreads to neighbouring pixels and is counted as two lower-energy photons, worse with smaller pixels |
| K-escape | Fluorescence X-rays from the detector material escape, so energy is recorded too low |
| Cost and data | Expensive detectors and very large datasets that need fast reconstruction and storage |
| New image texture | Radiologists and protocols need time to adapt |
Where it is used first
- Cardiac: coronary plaque, stents and calcium scoring.
- Chest: interstitial lung disease and low-dose lung screening.
- MSK and head and neck: trabecular bone, temporal bone and urate versus calcium crystal differentiation.
- Oncology: iodine quantification to separate enhancement from calcification and to monitor treatment.
- Paediatrics: diagnostic images at lower dose.
Exam angle
For FRCR physics and similar exams, remember: direct conversion, energy thresholds remove electronic noise, no septa so better dose efficiency and resolution, spectral data from every scan, and the three detector problems (pile-up, charge sharing, K-escape). Test yourself with our free FRCR physics questions and read MRI physics made simple.
Frequently asked questions
What is photon-counting CT?
A CT scanner whose detectors count each X-ray photon individually and measure its energy, using direct-conversion semiconductors instead of scintillators. It gives higher resolution, less noise and spectral data from every scan.
Is photon-counting CT better than dual-energy CT?
It provides spectral information from every scan with more energy bins and better spatial resolution, without dual sources or kV switching. Dual-energy CT is cheaper and widely available, and both are clinically useful.
Does photon-counting CT reduce radiation dose?
It can, mainly because electronic noise is removed and detector efficiency improves, which helps most in low-dose and paediatric scans. The actual dose depends on the protocol.
What are the limitations of photon-counting detectors?
Pulse pile-up at high count rates, charge sharing between small pixels, K-escape from the detector material, high cost and very large datasets.
Sources: Articl: Principles of photon-counting CT; Willemink et al., Radiology 2018; GE HealthCare Photonova Spectra clearance news (March 2026).