Single quantum imaging detects and measures individual photons or particles rather than treating incoming radiation as a continuous signal. For industrial automation and inspection, the approach offers a way to extract more information from radiation-based images: not only where a quantum was detected, but also its energy and time of arrival. The central engineering challenge is deciding how much processing to place in each pixel, and how that capability can be retained as pixels shrink.
Research on the Medipix and Timepix detector families, developed for particle physics and adapted to applications including X-ray imaging, illustrates the trade-off. Larger pixels can accommodate complex signal-processing functions. Smaller pixels can improve spatial precision and reduce capacitance, but leave less room for circuitry. This balance matters when evaluating detectors for applications that need material information, high event rates or precise timing.
What makes single quantum imaging different?
In conventional imaging, a sensor often integrates light or radiation over time. A single-quantum detector instead registers individual events and can measure properties of each one. In visible-light imaging, detecting a single photon-generated electron without avalanche amplification requires a very small pixel capacitance. At higher energies, including X-rays, the quantum can generate many electron-hole pairs in a semiconductor sensor, producing a signal that can be measured electronically.
That signal can carry more than position information. Depending on the detector design, pixel electronics can record an event’s arrival time and estimate its energy from the signal amplitude. When energy is measured, the image can distinguish radiation events associated with different energy levels, potentially providing information about the material being inspected.
Achieving this requires more than a sensitive sensor. A useful imaging matrix needs high active-area coverage, reliable event detection and enough local or downstream processing to measure relevant event properties. The more functions assigned to each pixel, the more circuit area and design complexity are required.
From particle tracking to X-ray inspection
Hybrid pixel detectors used in particle physics combine a semiconductor sensor with a separate CMOS readout chip. In Large Hadron Collider experiments, these assemblies record particle tracks across large detector surfaces. Their pixels measure charge created in the sensor, while the readout electronics amplify and process the signal. Timing and temporary on-pixel storage help manage events before data are selected for readout.
The Medipix collaboration adapted aspects of this approach for imaging beyond particle physics, particularly for X-rays. The hybrid architecture allows a readout chip to be paired with different sensor materials, while parallel processing allows pixels to register events locally. Timepix designs add time and signal-amplitude measurement, including time-over-threshold methods that use the duration of a signal above a threshold to estimate its size.
These features are relevant wherever an image needs to convey more than a simple count. In energy-resolved X-ray imaging, for example, separating events by energy can support material differentiation. The source describes rust detection during steel-pipeline scanning and distinguishing tissue types in medical CT as examples of the information such measurements can provide.
In-pixel processing can improve event handling
When one quantum deposits charge across several neighboring pixels, reading each pixel independently can make it harder to estimate the original event’s energy. Medipix3 addresses this with Charge Summing Mode: signals from adjacent pixels are combined on-chip and assigned to the pixel with the largest signal. This helps recover the event’s total charge and improves the energy spectrum compared with single-pixel measurement.
Local processing can also reduce the need to send every raw pixel response elsewhere for correlation. That is important at high event rates, when data transmission and storage become practical constraints. Timepix3 and Timepix4 designs use comparator-driven transmission of pixel data, and the architecture supports precise event timing. In Timepix3, the reported time precision is 1.56 nanoseconds; the paper describes better than 200-picosecond precision as an expectation for Timepix4, rather than a demonstrated result.
Timing can reveal event structure within a sensor as well as at its surface. In sufficiently thick sensor materials, charge carriers take time to travel to the collection electrode. Measuring signal timing can therefore help infer where within the sensor an event originated and support three-dimensional reconstruction of particle tracks.
The pixel-size trade-off
Smaller pixels can improve position measurement and help distinguish nearby events. But shrinking a pixel also reduces the area available for amplification, timing, digitization, memory and data handling. A detector designer must therefore weigh spatial resolution against the processing needed to extract useful information from each event.
Very small, low-capacitance “attopixels” offer a potential route to detecting signals from individual electrons without avalanche gain. Yet small dimensions alone do not provide the sophisticated measurement functions needed for energy-resolved imaging or precise time stamping. The paper’s proposed direction is hierarchical processing: simple sensing elements share more complex functions across groups of pixels or in stacked chip layers.
Existing Timepix designs provide an early example of sharing. Groups of eight pixels can use a common voltage-controlled oscillator and memory resources. More extensive 3D CMOS integration could place additional processing in multiple chip layers while limiting the footprint at the pixel surface. This remains a development path, not a reason to assume that every small-pixel detector already offers complex per-event analysis.
Practical implications for imaging-system design
For engineers and technical decision-makers, the detector should be assessed against the information the application needs. Key questions include:
- Is event energy important? If material discrimination is a goal, energy measurement and charge sharing between pixels may matter as much as nominal pixel size.
- Is timing a requirement? Time stamping can support high-rate operation and, in some configurations, reconstruction of event paths within the sensor.
- Where should processing occur? In-pixel or on-chip processing can reduce the burden of correlating raw events remotely, but requires more capable electronics.
- What sensor and geometry are appropriate? Hybrid detectors allow different sensor materials and thicknesses, while sensor depth and pixel pitch affect signal generation and spatial measurement.
Single quantum imaging is therefore not simply a race toward smaller pixels. Its value comes from matching sensor material, pixel geometry and processing architecture to the measurement task. Larger intelligent pixels can provide room for event-level analysis; smaller pixels can offer finer spatial encoding, especially when processing is shared through hierarchical or 3D integration. For industrial inspection, the practical question is which combination delivers the needed information about position, energy and timing without exceeding system requirements for data handling.