Photon-Counting CT
A category of CT detector technology that counts and energy-resolves individual X-ray photons rather than integrating their summed signal as a continuous current — the most fundamental architectural change to CT detection in 30+ years. Conventional CT detectors are energy-integrating: photons strike a scintillator, the scintillator emits visible light, photodiodes convert the light to current, and the current is integrated over a sampling window. Spatial information about which photons arrived where is preserved, but energy information about each individual photon is lost — the detector outputs a single intensity value per pixel per sample.
A photon-counting detector replaces the scintillator + photodiode stack with a direct-conversion semiconductor (cadmium telluride, CdTe, on the Siemens NAEOTOM Alpha; cadmium-zinc-telluride, CZT, on research platforms). X-ray photons absorb in the semiconductor and produce a charge cloud directly proportional to the photon energy. Fast electronics count each charge event and bin it by energy. Every scan is inherently spectral; every photon contributes; electronic noise below a counting threshold is rejected.
Why buyers care
- Spectral imaging on every scan — material decomposition, virtual monoenergetic, iodine quantification, virtual non-contrast — without dedicated dual-energy acquisition workflow.
- Higher spatial resolution at the detector — direct-conversion eliminates scintillator-light-spread blur. NAEOTOM Alpha achieves ~0.2 mm at the detector, finest in production CT.
- Lower dose at equivalent image quality — improved geometric efficiency + electronic noise rejection enables ~15–30% dose reduction vs energy-integrating CT (vendor / literature claims; site-dependent).
- No spectral-acquisition workflow penalty — clinicians don't choose "do I want spectral data on this exam"; they always have it.
Why engineers care
- CdTe / CZT detector materials are different from scintillator-photodiode stacks — different failure modes, different aftermarket / refurb posture, different service ecosystem.
- Per-photon counting electronics run at extremely high speed (millions of counts per pixel per second); the ASIC architecture is custom and proprietary. Service depth is OEM-only at this point in the technology lifecycle.
- Calibration is more complex — energy-bin calibration adds a dimension to the calibration suite vs energy-integrating CT.
- Dose-output and detector efficiency trade-off shifts — photon-counting detectors are particularly sensitive at low energies, which interacts with kVp / filtration choices in protocols.
Current platforms
- Siemens NAEOTOM Alpha — first FDA-cleared photon-counting CT (September 2021). Dual-source CdTe configuration.
- NAEOTOM Alpha.Pro — current dual-source configuration with workflow refinements.
- NAEOTOM Alpha.Prime — first single-source photon-counting CT, lower-cost / lower-siting-burden variant.
Other OEMs have announced photon-counting CT development but have not (as of early 2026) shipped production units. Canon Aquilion ONE / PRISM offers spectral imaging via different physics and is not photon-counting; same applies to Philips IQon (dual-layer detector) and GE Revolution Apex (Gemstone Clarity scintillator).
Refurb posture
- No meaningful refurb market yet — installed base is too young.
- Service-contract continuity dominates operating economics; CdTe detector and counting-electronics maintenance are vendor-only.
- Technology is too new for a mature lifecycle picture — expectations for CdTe detector aging at 8 / 10 / 12 years post-install are not yet established by deployment data.