Scientists at the National Institute of Standards and Technology (NIST) are proposing a new method for calibrating computed tomography (CT) scanners that could yield a common standard aimed at improving the comparison of scans from different machines. This method will help establish the first CT measurement standards related to the International System of Units by creating a more precise definition of the units used in CT.

Better calibration of CT scanners could make diagnosis more efficient and less expensive, NIST scientists wrote in a paper published in the journal PLOS One. Interchangeable measurement systems for calibrating CT machines are lacking in the field of radiology, and medical institutions must take action on a regular basis. Calibration involves scanning an object called a phantom with a known radio density to check that its measurements give the appropriate number of Hounsfield units (HU). The radiation density of an object or its ability to block X-rays is measured in HU.

In order to establish a common standard of calibration among scanners, scientists must overcome some challenges to reconcile differences between CT machine manufacturers, the NIST researchers said.

One challenge is that the tubes of the CT scanner generate photon beams of different wavelengths, and the overall effect on the phantom must be averaged. Another complication is that the X-ray light of the tube varies depending on the type of scan. Dense body parts require more penetrating X-rays, so scanner operators must adjust tube voltages to match the job. This change in the beam’s spectrum makes it harder to ensure that the calibration is correct for all voltages, the NIST researchers said.

? ? Overcoming the uncertainties created by the tube’s broad X-ray spectrum and tube voltage settings, the NIST researchers filled several phantoms with different concentrations of powdered chemicals that are common in vivo, using CT compared to their the radiodensity. The comparison relates HU to moles per cubic meter, both of which are SI units. The NIST researchers say better comparisons between scanners could lead to identifying critical points for disease, such as assigning a specific Hounsfield score to emphysema. It also eliminates the need for a second imaging process.