Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 62220-1-2: Detective Quantum Efficiency of Digital X-Ray Imaging Devices for Mammography
A practical guide to DQE measurement for mammographic detectors — from radiation quality to noise power spectrum analysis
IEC 62220-1-2, published in 2007, is the international standard that defines a unified method for determining the Detective Quantum Efficiency (DQE) of digital X-ray imaging devices specifically used in mammography. Prepared by IEC Subcommittee 62B (Diagnostic imaging equipment), this standard addresses a critical need in medical physics: ensuring that DQE values reported by different manufacturers are measured under comparable conditions. The DQE is widely recognised as the most comprehensive single-parameter descriptor of an X-ray imaging detector’s performance, quantifying how efficiently the detector preserves the signal-to-noise ratio (SNR) from the incident radiation field through to the output digital image.
DQE is fundamentally a measure of dose efficiency. A higher DQE means that for the same image quality, a lower patient radiation dose can be used — directly impacting patient safety in screening mammography programmes.
Measurement Principle and Key Parameters
The DQE is defined as the ratio of the squared Modulation Transfer Function (MTF) to the Noise Power Spectrum (NPS), normalised by the incident photon fluence. In mathematical terms: DQE(u,v) = [MTF(u,v)]^2 / [W(u,v) x Q], where W(u,v) is the NPS of the output image and Q is the mean number of incident X-ray photons per unit area at the detector surface. This formulation elegantly combines the detector’s spatial resolution (MTF), noise performance (NPS), and sensitivity to radiation (Q) into a single frequency-dependent metric.
The standard specifies that the conversion function of the detector must first be measured to linearise the raw pixel data. The MTF is determined using the edge method (or slit method), and the NPS is computed from uniformly exposed images using Fourier transform techniques. Lag effects — the influence of previous exposures on subsequent images — must be quantified and corrected where necessary, as residual lag can artificially inflate DQE estimates.
| Parameter | Symbol | Measurement Method | Significance |
|---|---|---|---|
| Modulation Transfer Function | MTF(u,v) | Edge / slit method | Spatial resolution |
| Noise Power Spectrum | NPS / W(u,v) | Uniform exposure + 2D FFT | Noise texture and magnitude |
| Photon Fluence | Q | Air kerma measurement + spectrum | Dose normalisation |
| Conversion Function | DN vs. Q | Step-wedge exposure series | System linearisation |
| Lag Effect | — | Sequential exposure decay | Image memory correction |
Attention: When measuring the NPS, the standard requires a minimum of 4 uniformly exposed images, each containing at least 256 x 256 pixels of region of interest data. Inadequate region-of-interest size or an insufficient number of images will introduce significant statistical uncertainty in the low-frequency DQE region, where clinical diagnostic information is most critical.
Radiation Quality and Test Conditions for Mammography
Unlike general radiography, mammography uses lower X-ray tube voltages (25-35 kV) with molybdenum or rhodium anode targets and corresponding beam filters to optimise contrast for soft tissue imaging. IEC 62220-1-2 specifies the RQA-M radiation qualities as defined in IEC 61267, including Mo/Mo (RQA-M 2 as the primary quality), Mo/Rh, Rh/Rh, and W/Rh combinations. The standard nominal first half-value layer (HVL) values range from 0.56 mm Al to 0.74 mm Al depending on the radiation quality selected.
The test geometry requires that the anti-scatter grid be removed for DQE measurements, as the grid’s presence would modify both the incident spectrum and the detector’s measured MTF. The detector surface is defined as the accessible plane closest to the image receptor after removal of removable components. Air kerma at the detector surface is measured using calibrated radiation meters with an uncertainty (coverage factor k=2) of less than 5%.
Q1: Why is DQE measured without the anti-scatter grid?
A: The grid affects both the incident X-ray spectrum and the detector’s spatial resolution characteristics. By measuring DQE without the grid, the standard isolates the intrinsic performance of the detector itself, separate from the grid’s scatter rejection properties.
A: The grid affects both the incident X-ray spectrum and the detector’s spatial resolution characteristics. By measuring DQE without the grid, the standard isolates the intrinsic performance of the detector itself, separate from the grid’s scatter rejection properties.
Q2: How does DQE relate to patient dose in mammography?
A: Directly. A detector with DQE of 0.7 at zero spatial frequency requires approximately 30% less dose to achieve the same SNR as a detector with DQE of 0.5, assuming equal MTF. In screening programmes, this translates to a tangible reduction in population radiation risk.
A: Directly. A detector with DQE of 0.7 at zero spatial frequency requires approximately 30% less dose to achieve the same SNR as a detector with DQE of 0.5, assuming equal MTF. In screening programmes, this translates to a tangible reduction in population radiation risk.
Q3: What is the typical DQE of modern digital mammography detectors?
A: Modern amorphous selenium direct-conversion detectors achieve DQE(0) of 0.7-0.85, while indirect CsI:Tl-based flat-panel detectors achieve DQE(0) of 0.55-0.75. Photon-counting detectors can exceed DQE(0) of 0.85.
A: Modern amorphous selenium direct-conversion detectors achieve DQE(0) of 0.7-0.85, while indirect CsI:Tl-based flat-panel detectors achieve DQE(0) of 0.55-0.75. Photon-counting detectors can exceed DQE(0) of 0.85.
Q4: Can DQE be compared between different radiation qualities?
A: DQE is a function of both spatial frequency and the incident X-ray spectrum. Comparisons should only be made at the same radiation quality (e.g., RQA-M 2). Reporting DQE at multiple spectra is recommended for a complete characterisation.
A: DQE is a function of both spatial frequency and the incident X-ray spectrum. Comparisons should only be made at the same radiation quality (e.g., RQA-M 2). Reporting DQE at multiple spectra is recommended for a complete characterisation.
In practical engineering terms, the DQE measurement procedure requires careful attention to experimental conditions. The X-ray tube must be warmed up and stabilised before any measurements are taken. The standard requires that the percentage ripple of the high-voltage generator be equal to or less than 4%, and the nominal focal spot value must not exceed 0.4. These conditions ensure that the incident radiation field is sufficiently uniform and that the measured MTF is not degraded by focal spot blur. The edge test device used for MTF measurement must be fabricated from a material with high atomic number (typically tungsten or tantalum) and must be positioned at a slight angle (1.5-3 degrees) to the detector pixel matrix to enable oversampling of the edge spread function.
The clinical relevance of DQE extends beyond the laboratory. Regulatory authorities such as the FDA and national competent bodies require DQE conformance as part of the pre-market approval process for digital mammography systems. The standard conformance statement format ensures that reported DQE values are accompanied by sufficient contextual information including radiation quality, air kerma, detector temperature, and pixel pitch to allow meaningful comparison between systems. Mammography screening programmes that have transitioned from screen-film to digital detectors have seen a 20-30% reduction in average glandular dose per examination, largely attributable to the improved DQE of modern digital detectors.
The measurement of DQE is not merely an academic exercise — it has direct implications for clinical practice. In mammography, where the detection of microcalcifications (typically 100-500 micrometres in size) is critical for early breast cancer diagnosis, the combination of high spatial resolution and low noise is paramount. A detector with superior DQE can visualise these subtle features at lower radiation doses, directly benefiting patients through reduced cancer risk from screening radiation. The standard also emphasises that DQE measurements should be performed at multiple air kerma levels spanning the clinical operating range, ensuring that the detector performance is optimised across the full exposure range used in practice.
