The advent of iterative reconstruction (IR) algorithms has been a major breakthrough in the optimization of dose and image quality for computed tomography (CT) protocols in clinical routine [1,2]. These IR algorithms have led to a significant reduction in radiation doses for many clinical applications [3], [4], [5], [6], [7]. However, these algorithms have non-linear and non-stationary properties that make the spatial resolution dependent on contrast and noise conditions, and modify the image texture [8,9]. These properties render obsolete the use of classical metrics, which are based on simplifying assumptions of stationarity and linearity of the signal. It is necessary to reintroduce so-called advanced metrics to evaluate image quality such as the noise power spectrum (NPS), the task-based transfer function (TTF) and the detectability index (d’). The NPS is used to evaluate noise texture and noise magnitude in the frequency domain and the TTF to evaluate spatial resolution under conditions of contrast and noise, close to the lesions encountered in clinical practice. Last, the detectability index is used to estimate the radiologist's ability to detect a simulated lesion.

These metrics are very useful to evaluate the performance of reconstruction algorithms but also to optimize the doses of a given protocol or to evaluate and compare a new tool or technology. However, the use of these metrics remains complex because they are calculated in the frequency domain and require an appropriate calculation platform such as a software.

In 2017, there was no software available to calculate all these three metrics for all types of image quality phantom and CT images. The French Society of Medical Physics therefore set up a working group of 3 physicists to develop software to address this issue. The iQMetrix-CT software was developed from 2017 to 2022.

The purpose of this technical note was to explain how the new iQMetrix-CT software works, as well as to describe its current potential and discuss its future applications.