Optimizing X-ray imaging procedures

Bildoptimering vid röntgendiagnostik

In this field of research, we balance patient organ doses and image quality in order to minimise the doses while maintaining or improving the quality of the radiographs. 

Our challenge is to develop methods to define and measure clinical image quality so that the optimisation process can be efficient and generally applicable. We therefore use European radiographic image criteria and advanced computer models of the complete imaging system, including the patient and radiologist. It is anticipated that if successful, our developed methods can facilitate optimisation and improve radiation safety.

In radiology, the clinical staff is obliged to minimise the x-ray exposure, but at the same time make sure that the image quality is sufficient for a correct diagnosis. This process is called dose and image quality optimisation. Our research shows that large dose reductions are still possible without reducing clinical image quality. 

Our objective is to develop and use computer simulation software to model the complete x-ray imaging system to predict image quality and absorbed doses in the patient. To reach this goal, we are developing a so-called virtual x-ray system, by searching for correlations between the subjective image quality made by radiologists and of objective image quality by a so called model observer. The model observer computes a detectability index for a pathological task, for example the signal-to-noise ratio of a contrast-filled heart vessel or a chest or breast tumour. The signal-to-noise ratio measures how well this pathology can be detected by the radiologist in the images where its visibility is limited by quantum noise. 

The virtual X-ray system provides unique, cost-efficient possibilities, not just for evaluating existing X-ray systems, but also for exploring future imaging systems before constructing expensive prototype systems. The research therefore gives important design information to manufacturers of new X-ray imaging systems.

An example of a model observer is given in the graph below. Here, the dose efficiency (signal-to-noise ratio per mean absorbed dose [µGy-1]) is given for three different imaging techniques in an angiography examination. The three different techniques (the three curves in the graph) all indicate that the dose efficiency peaks at a tube potential of 60 kV, which is much less than what is commonly used today (80 kV). Our results show clear indications that dose reductions up to 50% are possible using optimal settings in clinical practice for the benefit of the patient.

The figure illustrates how our virtual X-ray system was validated, as the measured (single marker points) and calculated (lines) data both coincide and indicate that dose efficiency peaks at a tube potential of 60 kV.

The figure illustrates how our virtual X-ray system was validated, as the measured (single marker points) and calculated (lines) data both coincide and indicate that dose efficiency peaks at a tube potential of 60 kV.

Research Group
Show/Hide content

Research Area
Show/Hide content

Research Center
Show/Hide content

Organisation
Show/Hide content