Thales's Digital Radiography Technology: enhancing X-ray imaging quality

Healthcare professionals are under immense pressure to deliver high quality care, often in difficult circumstances. To meet these demands, our customers are in need of the best technologies that generate optimised image quality whilst operating at a lower dose. New and ever-developing technology in the radiography industry has has quickly advanced improving the accuracy of the diagnosis while generating debates among the major contributors.

One recent debate concerns pixel pitch and the vital role it plays in purchasing decisions. What exactly is it, why is it important, and does an "optimum" level actually exist?

Pixel pitch is the distance between the centers of two consecutive photon-detecting pixels. It is measured in micrometers (µm) and generally ranges between 75 and 200µm. To better understand the importance of pixel pitch, it is necessary to distinguish between detectivity and resolution.

The most important characteristic of performance is detectivity, allowing the detection of a tumour or fracture. Detectivity is an expression of the SNR (Signal/Noise Ratio), which is the ratio of the signal (average amount of X-rays which arrive at the detector)  to the noise (the visual variations in this average). One of the ways to improve the SNR is to improve the signal by increasing the pixel pitch (or size of the pixel). This will decrease the number of pixels, and therefore increase the number of X-ray photons detected by the pixel.

Another very important parameter linked to detectivity is the DQE (Detective Quantum Efficiency), which is the ratio of SNR² (number of photons visible on the detector) to the dose. It measures the proportion of information carried by the X-rays that is visible on the screen. As the dose decreases, the number of photons visible on the screen (SNR²) decreases proportionally, meaning that the DQE is not affected. While there may seem to be more noise, there is no loss of information visible in the image.

The second major characteristic influencing performance is the MTF (Modular Transfer Function), which is the resolution of the image at a given spatial frequency. The resolution is ultimately the maximum number of line pairs per millimeter (lp/mm) visible. With a low MTF, the shape would be blurred but with a high MTF it would be clearly defined and easily visible. Detectivity is a matter of the signal/noise ratio, whereas resolution is a matter of the pixel size of the detector.

The graphic below compares one image (left) with a high resolution (MTF) but low detectivity (DQE) and a second image (right) with high DQE but low MTF. On the left, the shape is well defined thanks to the high resolution but with more noise. On the right, the stars are bright and clearly visible, but it would be impossible to measure their shape or diameter as the resolution is not high enough. A high-resolution image is not, in many cases, very useful to radiologists, as their priority is first to localise the potential pathology (tumour or fracture), and not to measure the exact shape of very small pathologies. As detectivity comes before resolution, the pixel size of the detector can be larger without losing any information. 

Is there an "optimum" pixel pitch?

A larger pixel pitch means a higher SNR, higher detectivity, and an image which is easier to read. As pointed out earlier, it is detectivity which is of main concern to radiologists and this is therefore a key argument for large pixel pitch.

That being said, resolution is still important and so there is a certain level that is set by radiologists as a minimum requirement, as is shown in the diagram above.  Three independent French radiologists recently carried out a study in arthritis detection, where very tiny bone structures were detected and counted. These structures are much smaller than 100µm, so one would expect that having a pixel pitch lower than 100µm would help. The exercise was performed, with 50 and 150µm pitch detectors, on a large set of radiographies. The result of the study showed that the same number of structures were systematically counted whatever the pixel pitch. Small pixels didn’t help to see more structures, resolution did not help, and detectivity was the most important feature.

Responding to the needs of healthcare professionals

In response to the debate and questions posed, the combination of Thales’s experience and Trixell’s state-of-the-art technology proposes the Pixium Portable 3543 DR, with an optimum pixel pitch of 160µm achieving a maximised DQE & MTF for the range of 1-2 lp/mm, in the diagnostic area. The Pixium® Portable 3543 DR facilitates the daily work of medical professionals, improving the quality of images and increasing the pace and accuracy of patient diagnosis.

Click here to find out more about our range of Radiology technologies!