Dr Susan Astley
Research
Automatic detection of mammographic abnormalities.
We have investigated the use of Bayesian statistics to improve specificity in the detection of microcalcifications, which are one of the earliest signs of cancer, and found that combining different cues was effective in improving detection (Astley & Taylor 90). The detection of asymmetry is is more difficult, as the breasts are variable in appearance and differ naturally. The technique we have developed is based on identifying anatomically similar regions using the transportation algorithm to measure similarity (Miller and Astley 93, Board et al 04). We have also investigated the detection of spiculated masses and distortion (Zwiggelaar et al 04). One of the key ideas behind this approach was model-based classification of linear structures in the digitised images (Parr et al 96).
Computer-Aided Detection (CAD)
Prompts based on algorithms are presented to radiologists to attract attention to potential abnormalities. We have used synthetic, mammogram-like images as a basis for psychophysical experiments with the objective of determining how accurate prompting algorithms must be to achieve a significant improvement in human detection performance (Hutt et al 94, Astley et al 98). This work demonstrated the importance of specificity; however, the situation is complex, particularly when multiple algorithms are used, and this remains an active area of interest. We have evaluated three commercially available prompting systems and conducted both retrospective and prospective clinical trials to compare single reading with CAD and the current standard clinical practice of double reading (Astley et al 02, Gilbert et al 06).
Breast Density and Risk
One problem in mammographic analysis is the calibration of pixel values in images from one mammogram to the next, which is essential if measures of change in density or pattern associated with increased risk are to be made. We have measured the bending of the compression plate by imaging markers on the plate (Diffey et al 2008), and designed and fabricated a calibration object suitable for clinical use (Smith et al 96, Diffey et al 2006). We have used this technique to measure change in women enrolled in a ‘Lifestyle’ study at the Christie Hospital; results have shown that as women lose weight, the volume of fat in their breasts decreases significantly, whilst there is only a small change in the volume of gland (Patel et al 2006). This is important because current estimates of breast density (a known risk factor for breast cancer) are based on the percentage of dense tissue in the breast. Thus as a woman loses weight, percentage density would rise indicating an increase in risk, when in fact increased weight itself is a risk factor. Our method computes the volumes of dense and fatty breast tissue independently, so a more accurate method of assessing risk can be developed. We are currently undertaking a pilot study of 3,000 women attending for breast screening, with the aim of establishing the distribution of fatty and glandular volumes in the normal screening population and relating this to age and other known risk factors (Diffey, Maxwell et al 08). This is expected to provide a springboard for other population based studies looking at individual risk assessment in the screening programme.
Lesion Modeling
We have developed a method for generating realistic synthetic masses by statistically modelling spiculated breast lesions. Many of the synthetic lesions are indistinguishable from real lesions by consultant breast radiologists (Caulkin and Astley 00, Berks et al 08a). We are currently looking at the relationship of the lesions to normal breast tissue (Berks et al 08b).
Teaching
MMS Lead for Intercalated Degrees
Portfolio Tutor (years 1 and 2)
Seminars on Research for MSc students
Collaborators and affiliated staff
Dr Caroline Boggis, Nightingale Breast Centre, Manchester
Dr Alan Hufton, North West Medical Physics, Christie Hospital, Manchester
Prof Fiona Gilbert, University of Aberdeen
Prof Michael Brady, University of Oxford
Publications
2011
- Berks M, Chen Z, Tresadern P, Astley S, Taylor C. (2011). Detecting and classifying linear structures in mammograms using random forests. Presented at Information Processing in Medical Imaging. eScholarID:153420
- Makaronidis J; Berks M; Sergeant J, Morris J, Boggis CRM, Wilson M, Barr N, Astley S. (2011). Assessment of breast density: reader performance using synthetic mammographic images. Presented at SPIE Medical Imaging. Lake Buena Vista, Florida: SPIE. eScholarID:153424
- R. Sadeghzadeh; M. Berks; S. M. Astley; C. J. Taylor. (2011). Evaluation of blood vessel detection methods. Presented at SPIE Medical Imaging. Lake Buena Vista, Florida, USA: SPIE. eScholarID:153433
- Brentnall AR, Duffy SW, Crowder MJ, Gillan MG, Astley SM, Wallis MG, James J, Boggis Cr, Gilbert FJ. (2011). A method for exploratory repeated-measures analysis applied to a breast-cancer screening study. The Annals of Applied Statistics, 5(4), 2448-2469. eScholarID:153422 | DOI:10.1214/11-AOAS481
2010
- 10. J Diffey, J Morrison, M Berks, J Greene, J Morris, A Maxwell, C Boggis, A Hufton, and S Astley. (2010). Volumetric Breast Density and Breast Cancer Risk Factors in a Screening Population. Presented at International Workshop on Digital Mammography. Girona, Spain: Springer. eScholarID:153447
- 4. C Jeffries Chung, J Diffey, J Morrison, M Berks, R Verow, J Morris, M Wilson, C Boggis, N Barr, SM Astley. (2010). Automated Assessment of Area of Dense Tissue in the Breast: A comparison with Human Estimation. Presented at International Workshop on Digital Mammography. Girona: Springer. eScholarID:153438
- Jenny Diffey, Michael Berks, Alan Hufton, Camilla Chung, Rosanne Verow, Joanna Morrison, Mary Wilson, Caroline Boggis, Julie Morris, Anthony Maxwell and Susan Astley. (2010). A stepwedge-based method for measuring breast density: observer variability and comparison with human reading. Presented at SPIE Medical Imaging. eScholarID:153429
- M Berks, C Taylor, R Rahim, C Boggis, S Astley. (2010). Modelling structural deformations in mammographic tissue using the dual tree complex wavelet. Presented at International Workshop on Digital Mammography. Girona, Spain: Springer. eScholarID:153443
- M Berks, C Taylor, R Rahim, D Barbarosa da Silva, C Boggis and S Astley. (2010). Synthesising malignant breast masses in normal mammograms. Presented at International Workshop on Digital Mammography. Girona, Spain: Springer. eScholarID:153444
- Michael Berks; David Barbosa da Silva; Caroline Boggis; Sue Astley. (2010). Evaluating the realism of synthetically generated mammographic lesions: an observer study. Presented at SPIE Medical Imaging 2010. SPIE. eScholarID:153435
- N Barr, C Boggis, N Barr, M Wilson, J Morris, M Berks and S Astley. (2010). Comparison of microcalcification detection rates and recall rates in digital and analogue mammography. Presented at International Workshop on Digital Mammography. Girona, Spain: Springer. eScholarID:153445
- R Verow, M Berks, J Diffey, C Chung, J Morrison, M Wilson, C Boggis, N Barr, J Morris, A Hufton and S Astley. (2010). Inter and Intra Observer Variability in a Semi-Automatic Method for Measuring Volumetric Breast Density. Presented at International Workshop on Digital Mammography. Girona, Spain: Springer. eScholarID:153446
- Sukha, M Berks, J Morris, C Boggis, M Wilson, N Barr, SM Astley. (2010). Visual Assessment of Breast Density in Digital Mammograms. Presented at International Workshoip on Digital Mammography. Girona, Spain: Springer. eScholarID:153441
- Jonathan J. James,Fiona J. Gilbert,Matthew G. Wallis ,Maureen G. C. Gillan, Susan M. Astley ,Caroline R. M. Boggis ,Olorunsola F. Agbaje, Adam R. Brentnall, Stephen W. Duffy. (2010). Mammographic Features of Breast Cancers at Single Reading with Computer-aided Detection and at Double Reading in a Large Multicenter Prospective Trial of Computer-aided Detection: CADET II. Radiology, 256, 379-386. eScholarID:153427 | DOI:10.1148/radiol.10091899
- F.J..Gilbert, K.C.Young, S.M. Astley. P Whelehan, M G C Gillan. (2010). Digital Breast Tomosynthesis. NHS Breast Screening Programme. eScholarID:153428
2008
- Duffy SW, Nagtegaal ID, Astley SM, Gillan MGC, Mcgee MA, Boggis CRM, Wilson M, Beetles U, Griffiths M, Jain AK, Johnson J, Roberts R, Deans H, Duncan K Iyengar G, Griffiths P, Warwick J, Cuzick J, Gilbert FJ. (2008). Visually Assessed Breast Density: the need for two views. Breast Cancer Research, 10(R64), eScholarID:53169 | DOI:10.1186/bcr2123
- FJ Gilbert, SM Astley, MC Gillan, OF Agbaje, MG Wallis, J James, CRM Boggis, SW Duffy. (2008). A comparison of single reading with Computer Aided Detection and Double Reading for Screening Mammography. New England Journal of Medicine, 359, 1675-1684. eScholarID:53036 | DOI:10.1056/NEJMoa0803545
- Gilbert FJ, Astley SM, Boggis CRM, McGee MA, Griffiths PM, Duffy SW, Agbaje OF, Gillan MGC, Wilson M, Jain AK, Barr N, Beetles U, Griffiths MA, Johnson J, Roberts R, Deans H, Duncan KA, Iyengar G. (2008). Variable size CAD prompts and mammography film reader decisions. Breast Cancer Research, 10(R72), eScholarID:53115 | DOI:10.1186/bcr2137
2006
- ASTLEY SM; GILBERT F; BOGGIS C; DUFFY S; . (2006). Single reading with computer-aided detection and double reading of screening mammograms in the United Kingdom national breast screening programme. Radiology, 241(1), 47-53. eScholarID:1a10115 | DOI:10.1148/radiol.2411051092
2005
- ASTLEY SM. (2005). Evaluation of computer aided detection (CAD) prompting techniques for mammography. British Journal of Radiology, 78(special issue 1), S20-S25. eScholarID:1a10114 | DOI:10.1259/bjr/37221979
2004
- ASTLEY SM; GILBERT F. (2004). Computer-aided detection in mammography. Clinical Radiology, 59(5), 390-399. eScholarID:1a10117 | DOI:10.1016/j.crad.2003.11.017
- ASTLEY SM; ZWIGGELAAR R; BOGGIS C; TAYLOR C. (2004). Linear structures in mammographic images: detection and classification. IEEE Transactions on Medical Imaging, 23(9), 1077-1086. eScholarID:1a10116 | DOI:10.1109/TMI.2004.828675