KAHIBARO
18.1. What Are Singles?
Table of Contents
Detector events
In GATE, a single is the basic detected event in a radiation detector after the raw particle interactions have been processed by the digitizer chain. It represents one detection in one detector channel at one time, with one energy value and one position associated to it.
From the physics point of view, many particles can interact inside a detector volume and create multiple hits in one or several crystals. These hits correspond to individual energy depositions at specific positions and times. The digitizer then groups and processes these hits to decide whether they form a valid detector event. When this decision is positive, the result is stored as a single.
You can think of singles as the Monte Carlo representation of what a real acquisition system would call count events in a detector. Each single is close to what a pulse height analyzer or a data acquisition system would record after all analog and digital electronics have done their work. For example, in a PET scanner, a gamma photon that deposits energy in one LYSO crystal and passes all thresholds will usually appear as one single corresponding to that crystal.
In a typical GATE PET or SPECT simulation, singles are at the center of most data analysis. They are used to build energy spectra, to measure count rates per detector, and to form coincidences by pairing singles that occur within a coincidence time window. Later chapters will describe in detail how singles are produced from hits and how they are used to form coincidences, but here it is important to understand that singles correspond to detector level events, not to microscopic particle steps.
Each single usually carries a compact set of information. Common attributes include the deposited energy in the detector channel, the time of the event, and one or several detector identifiers, such as crystal ID or module ID. Some digitizer configurations can also store additional information like event IDs that link back to the original simulated history. The exact content of a single depends on the digitizer configuration defined in your simulation script.
A single represents one detector count after processing, not one physical interaction. It results from combining and filtering many underlying hits according to the digitizer settings.
Understanding this distinction avoids confusion when comparing the number of hits and the number of singles in your output. Often, there are many more hits than singles because several hits are merged into one detector event, and many low energy interactions may be discarded by thresholds.
Processed detector signals
The key idea behind singles is that they are already processed signals. The digitizer takes the raw hit collection and applies a sequence of operations that mimic the detector electronics. These operations can include energy summation, energy and time blurring, application of energy thresholds, mapping to readout channels, and possibly time clustering. The result of this full processing chain is a set of singles that approximates what a real detector system would output as digital events.
At an abstract level, the transformation can be summarized as:
$$\text{hits} \rightarrow \text{detector processing} \rightarrow \text{singles}$$
In this context, singles are not just a simplification of hits, they are a different representation that is closer to experimentally measured data. While hits contain detailed information about every microscopic interaction, singles store what would be recorded on disk in a real experiment. This is why singles are usually the starting point for imaging or dosimetry performance analysis.
Because singles represent processed signals, they already include the effects of detector imperfections that you have modeled in your digitizer chain. Energy blurring introduces finite energy resolution, time blurring introduces finite timing resolution, and energy windows remove events outside the chosen photopeak region. All these effects are reflected in the singles. For example, the shape and width of the 511 keV photopeak in a PET energy spectrum is determined only after hits have been blurred and thresholded to create singles.
The table below summarizes the conceptual difference between hits and singles.
| Property | Hits | Singles |
|---|---|---|
| Level of description | Physical interactions in materials | Processed detector signals |
| Typical number per photon | Many | Zero or one per detector channel |
| Includes electronics model | No, pure physics output | Yes, via digitizer chain |
| Typical use | Detailed detector studies, debugging | Imaging performance, coincidence building |
Singles are also the basis for coincidence sorting in PET simulations. Two singles detected in different crystals within a given time window can be paired into one coincidence event. This pairing does not use hits directly, only the time and position information stored in the singles. As a result, the statistical properties of coincidences, such as random and scatter fractions, depend strongly on how singles are produced and filtered.
When you analyze PET or SPECT performance in GATE, you almost always work with singles or coincidences, not raw hits. Singles are the bridge between detailed Monte Carlo physics and realistic detector output.
For beginners, an effective way to gain intuition is to run a simple simulation, save both hits and singles, and compare their distributions. You will see how many physical interactions are compressed into a much smaller, experimentally meaningful set of processed detector events.
Views: 11