Professor Peter Hogg
Professor of Radiography, Director and Research Dean at the School of Health and Society, University of Salford
Dr Andrew England
Senior Lecturer in Radiography, University of Salford
The radiation dose for any imaging examination is a serious concern and can be a limitation when attempting to preserve the diagnostic quality of the image.
We investigate and solve a range of clinically relevant imaging problems that practitioners face on a daily basis. Our research covers a range of imaging modalities and examinations. We use a range of methods within this research, including estimations of dose, modelling and measurement; image quality evaluations using physical and psychophysical measures; and observer performance assessment.
Avoiding false diagnoses from blurred images
Examples of this research include identification of causation for, and minimisation of reasons leading to, blurred full field digital mammography (FFDM) images. Blurred FFDM images can lead to false diagnoses and unnecessary repeat examinations. Repeat imaging can heighten client and family anxiety, increase overall breast screening costs and contribute to unnecessary radiation dosage.
Lifetime radiation risk prediction models
New mathematical models are being developed that aim to predict total lifetime radiation risk from screening mammography. These models are important as they can help clients make informed decisions about whether or not they wish to participate in screening. They also allow for comparisons between country-based screening programmes, intra-country screening regimes and FFDM machines.
Improving computer screen display to avoid inaccuracies
Identification of FFDM machine inaccuracies can lead to errors in practice, for instance inaccurate breast cancer risk classification for future screening and lesion localisation as part of biopsy processes. We are analysing computer screen display capabilities for technical checking and reporting on FFDM images. Our work here has already found that the technical checking monitors located in the clinical rooms might not be of an adequate standard to check images before sending them for reporting.
We are also working to improve detection of interval cancers, which occur between screening rounds; we are working to improve FFDM positioning technique to ensure standardisation (and thus improve accuracy), minimise discomfort and improve lesion detection performance. We’re also assessing the impact of visual acuity in breast cancer detection performance with a view to setting standards for eyesight checking.
Our practice-based research mainly relates to improving FFDM image quality and screening client experience during the imaging process. Examples of our research include identification and minimisation of compression force variability between and within practitioners. We were the first to prove the existence of this, initially focusing our work on a UK population.
Since then, we have published a major piece of research that assessed the entire screening population of Norway. Impact from this has been high, with an international software company introducing an automated method to assess practitioner variability. This, it is hoped, will allow for easy identification of outliers and the ability to implement solutions on a global scale.
Many aspects of this research have a close relationship with student learning on our undergraduate and postgraduate programmes, particularly the MSc Advanced Medical Imaging.
