Dose efficiency of CT systems: an automated procedure for evaluating dose and image quality

Experimental setup for determining dose efficiency

Background

Computed tomography (CT) plays a vital role in medical diagnosis but is associated with a relatively high radiation dose in comparison with a simple X-ray image. In terms of numbers, CT scans make up a relatively small proportion of examinations using X-rays – but they represent the greatest contribution to the population’s collective dose from X-ray applications.

In recent years, various advances in instrument engineering have enabled significant reductions in the radiation dose used for each examination – while producing images of comparable diagnostic quality. According to the principle of dose optimisation, diagnostic processes can only expose people to as much radiation as is necessary to produce an image of sufficient quality for diagnosis. When it comes to acquiring new CT systems, however, it is difficult for buyers at hospitals or medical practices to know which devices deliver what image quality per radiation dose – in other words, what their "dose efficiency" is like.

Objective

A research project in collaboration with the Federal Institute of Physics and Metrology (PTB) is intended to develop a standardised process for assessing the dose efficiency of CT systems. In the future, it will ideally be possible to use this technique in type testing.

Methods

Objects with low contrast represent a challenge to CT diagnostics. Researchers chose a suitable phantom with low-contrast structures with small diameters and a representative abdomen protocol. The generated image data is evaluated by a model-based observer, and the results are used to determine an objective parameter (the dose efficiency). Image data is evaluated and dose efficiency is determined as automatically as possible in the specially developed software.

Reproducibility was verified by multiple measurement series – first with the same device and the same phantom and then with various phantoms and the same device.

Implementation

So far, measurements have been carried out on 12 devices at various locations. These include devices from four manufacturers and 10 different types of devices, including systems with 16 to 192 detector rows and with one or two X-ray tubes. Further measurements are planned.

Previous results

Measurements to verify reproducibility delivered good results. Examinations with various phantoms revealed significant differences in some cases, and these differences must be taken into account with regard to type testing.

Outlook

There are plans to continue the measurement campaign and develop the software further (including the user interface).

Publications

Racine, D., Viry, A., Edyvean , S., & Verdun, F. R. (2017). Konzepte zur Charakterisierung klinischer CT Systeme unter Einbeziehung von Bildqualität und Dosis Vorhaben 3613S20007

Göppel, M., Anton, M., Gala, HH., Giussani, A., Trinkl, S., Renger, B., & Brix, G. Dose-efficiency quantification of computed tomography systems using a model-observer. Med Phys. 2023; 1- 12.