Objectives

1. Consensus on objective physical measures in screening techniques
2. Consensus and collation of dose surveys for new generation screening equipment
3. Consensus on equipment Surveys for new generation screening equipment
4. Consensus on optimisation process for new screening equipment

Description of work

1. Review of image quality approaches
2. Collation of surveys data on dose and functional performance
3. Collation of results of optimisation studies

Deliverables

1. Evaluation of dose quantities.
2. Collation of surveys and consensus approaches for special screening techniques including mammography, DEXA and paediatric radiology.
3. Physical assessment, optimisation and post processing for special screening and digital techniques.
   

Workpackage description

Whilst Detective Quantum Efficiency is generally accepted as the ultimate objective measure of image quality, it includes aspects of spatial resolution and noise for a delivered amount of X-ray quanta. It is proposed to extend this work, as several aspects need further attention and refinement.

It is proposed to undertake a survey to calculate the DQE for all available detector systems on the market by evaluating available data obtained. This includes slot scanning systems, for which a dedicated procedure will have to be established, and digital mammography systems for interventional work. A major difficulty for a routine assessment of the DQE is the measurement of the detector dose as it is not always possible to insert a dosemeter at the appropriate place. Moreover, solid state dosemeters may not be calibrated for this purpose. It is proposed to include aspects of spatial resolution and noise into a new quantity that relates these characteristics to entrance dose which is easy to measure in the field, rather than detector dose. Results obtained from the reviews of classical DQE assessment and the new quantity will be compared.

A method to assess the MTF for all detector positions to detect local problems of spatial resolution will be further automated and validated on different systems with resolution problems or in homogeneities.

Optimization of digital mammography systems includes many steps. Post processing and viewing conditions are recognized as being very critical components of the total digital mammography system. Previous approaches were based on CDMAM analysis the use of a phantom with small egg shells that simulate microcalcifications, and fully simulated lesions. This last tool is a new and particularly powerful approach to compare and optimize post processing settings and the look-up table of the monitors. It allows comparing the visibility of lesions on different systems. It is intended to study all the remaining problems for today. This includes:

1. The automation of a software tool that simulates lesions into raw data
2. Comparing optimisation the results obtained from approaches based upon signal to noise ratio or contrast to noise ratio constant for different breast thicknesses. This requires the calculation of the doses, in particular for the very dense and thin breasts, and the detectability of lesions for the different conditions. This analysis could be based on CDMAM readings for different conditions, signal to noise and contrast to noise calculations of particular inserts into test objects and using simulated lesions that are corrected for the different thicknesses under study.
3. A study to compare post processing tools from different vendors, using simulated lesions. Observer studies can be performed using real observers or mathematical models. The correlation between both methods will be established.
4. The use of simulated lesions allows the studying of improvements in visibility of microcalcifications that could be achieved if a detector had another MTF or another pixel size. These studies will have a large impact in the discussion regarding the minimal standards for a digital mammography system.

Due to the wide dynamical range of digital mammography systems, the detector can be used with wide range in dose that may go largely undetected. Dose monitoring in digital mammography is important for two reasons: the doses may have important consequences regarding the benefit in screening mammography and it may be used as an indirect way to assess the technical stability of the systems. It is of the utmost importance, in particular in a screening organisation, that images are delivered with a certain quality that is constant over time and at the lowest possible dose.

The problem in digital mammography when compared to another procedures covered in SENTINEL is that doses can be expressed in very different ways (e.g. entrance dose with or without backscatter, mean glandular dose, detector dose). In a first phase, the relative value, the ease of data acquisition and the combined use and role of these quantities in the radiological practice will be studied. The applicability to patient dose monitoring and assessment of detector stability will be tested.

For these very new systems, automated dose retrieval should be established. The necessary parameters to be collected will be listed and these data will be communicated to the industry. Improved software approaches will be advised. Mean glandular doses will be calculated from a larger number of images in a series of centres using different digital mammography systems for a range of technique factors as part of an optimisation process.

Patient doses will be linked to image quality indicators of the mammographic images acquired with these systems and compared to the average doses acquired with conventional systems.

A complete new generation of bone mineral densitometry scanners have been widely available and is very popular. Whilst there is no formal screening, the level of public awareness and interest amongst women's groups is such that it is now essential that studies regarding its standardisation and optimisation are undertaken. There is a potential for self referral on a regular basis. Clinical referral criteria will need to be established by expert groups. As their equipment tends to be used outside of traditional radiological settings there is a need for training and understanding of safety issues. Paediatric radiology approaches to patient dosimetry will be studied, in relation to population screening. This will be used to propose methods of assessing reference dose levels, image quality criteria and surveys of functional performance.