Anatomy of a Perfect Image: 3 Important Components of Image Quality

What affects image quality in flat panel detectors?
1. Pixel Pitch
Optimize the pixel pitch! (Remember, Pixel pitch describes the diameter of round pixels or the distance between dots <pixels> on a display screen. In digital radiography, the optimum pixel pitch is between 100-150 microns). A pixel pitch that is too small will give you blurred edge detail in the image anatomy, and also require a higher dose x-rays. A pixel pitch that is too high will produce a grainy image with poor spatial resolution or edge detail.

2. Scintillator

A scintillator is the part of the panel that converts x-radiation to light. A scintillator is made of crystals in two varieties:

  • Gadolinium-oxide sulfide (GadOX): unstructured crystal, lower cost, lower resolution, higher x-ray dose.
  • Cesium Iodide (Csi): structured crystal, higher cost, higher resolution, lower x-ray dose.
The difference between unstructured vs. structured scintillators can be likened to incandescent and LED light bulbs. Csi panels, like the LED light, is bright, crisp, and distinct, whereas GadOx, like incandescent bulbs, create a dimmer, more diffuse glare.

3. Electronics

Extremely low noise is ideal for electronics. What is noise? Noise is a random fluctuation in an electrical signal. There are a variety of causes for noise including the quality of the electrical components themselves, temperature, and conductance fluctuations. It is also important to consider battery life and efficiency of electronics. Some panels have an ideal pixel pitch, but must be charged every few hours and are hot to the touch because of inefficient electronics.

Bonus! Image Processing in Digital Radiography:
Lossy vs. Lossless Image Compression
Lossy image processing sacrifices image data to enhance edges. It allows for smaller file sizes by accepting some data loss either in the form of dropping bits per pixel or counting every other pixel.

This process detects white edges and then darkens the outside of the white edge with a “smudge” that look like dark halos around bone edges or tissue that are not real.

Lossless image processing never sacrifices image data to create a viewing effect. It issues a more sophisticated instruction as to location and value of each pixel. The instruction takes up less space than each individual pixel location and value. Bones are distinct and detail surrounding soft tissue accurately reflects the true density and shape of the anatomy.