This research was undertaken at the UK National Co-ordinating Centre for the Physics of Mammography (NCCPM), based at the Royal Surrey County Hospital in Guildford. Research projects included developing tools for the NHS Breast Screening Programme to allow various aspects of image quality in clinical films to be quantitatively assessed, and an investigation of false negative interval cancers. As well as details about my research projects you can also find links to other mammography pages via this website.
1. Quantitative assessment of image quality in mammograms
A methodology has been developed using film digitisation and image analysis, which allows the requirements of film contrast, exposure, and latitude in screening mammography to be quantitatively assessed. By using these image quality parameters the implications of the reduced latitude of modern high contrast films can be quantified and modifications to imaging procedures suggested to optimise the quality of the mammograms for all breast types. The results of the pilot study were presented at Symposium Mammographicum ‘98 as an educational poster paper and a scientific paper covering this work has been published in the BJR (abstract). A second educational poster paper was presented at ECR ‘99 showing how exposure data were used to estimate mammography film dynamic range requirements.
By optimising film screen mammography equipment and imaging procedures it may be possible to improve the detection of cancers. To facilitate this the most appropriate films and screens should be used. It is therefore necessary to assess the implications of using different film screen combinations. The latest very high contrast films from Kodak (min-R 2000) and Fuji (AD-M) have been assessed. An educational poster paper entitled "Optical density and latitude requirements of high contrast mammography films" was presented at ECR 2000 giving details of the study results. A scientific paper covering this work has been published in the BJR (abstract).
In film screen mammography systems the exposure is normally controlled by an automatic exposure control system (AEC). Achieving optimal film exposure is dependent on the correct performance of the AEC. A scientific paper has been published in the BJR detailing a "procedure for quantitatively assessing automatic exposure control in mammography" using clinical films (abstract).
The problems associated with imaging both the skin edge region and the skin line on mammograms was investigated. Details about quantitative and qualitative methods of assessment together with results from tests on a new film to cope with imaging both the skin edge and glandular tissues in mammograms satisfactorily, were the subject of a paper published in the BJR (abstract). Figure 1from this paper appeared on the front cover of the printed journal.
In the longer term the aim is to use some of the parameters that distinguish a good mammogram from a poor mammogram together with the data collected to develop a model of mammogram production. It will then be possible to analyse the effects of changes to imaging equipment and procedure ex vivo. Then as new films are introduced and different kVs, filters, etc are suggested it will be possible to determine the consequences for mammogram production without any women being exposed to X-rays.
2. Image features of true positive and false negative cancers in screening mammograms
The basic tools necessary for quantifying some important aspects of image quality in clinical mammograms have been developed and described by Meeson et al (1999, 2001) and in NHSBSP Report 99/2. These make it possible to evaluate quality measures of individual mammograms using a quantitative rather than a subjective approach. The aim of this work was to study the location, tissue background and imaging characteristics (including optical density) of true positive and false negative screens of breast cancers. This data can then aid in optimising the display of mammographic information with the objective of minimising false negative screens. The quantitative tools described in Report 99/2 and those developed specifically for this work were used in conjunction with the visual gradings of the radiologists to determine how the occurrence of interval cancers compared with screen detected cancers.
A scientific paper covering the results of this study has been published in the BJR (abstract). Key findings from this study were also presented as a detailed educational poster paper at Symposium Mammographicum 2002 and discussed in NHSBSP Report 01/07.
The two projects discussed above were funded by the National Co-ordinating Office of the NHS Breast Screening Programme in the UK. To make enquiries about the work presented here please e-mail me.
Last modified 25th February 2014