Role of automated level 2 grading in diabetic retinopathy screening

The role of automated level 2 grading within diabetic retinopathy screening collaborative network

Data was collected from five Scottish grading centres to ensure that the software developed in a prior study is generalisable to other parts of Scotland. Building on existing techniques, new computer algorithms were developed to detect referable retinopathy. Data was also gathered on health state values associated with correct and incorrect diagnosis to give an indication of the impact of false positive/negative results. An economic evaluation was also carried out to assess the costs and consequences of automated grading versus manual grading in the context of Scotland's national screening programme.

Outcome and Translation

The evidence from this study contributed to an independent review which concluded that automated grading should be implemented within the national diabetic retinopathy screening programme in Scotland (http://www.ndrs.scot.nhs.uk/ExecGrp/Docs/2009%2006%2009%20Retinal%20autograding%20Waugh%20FINAL.pdf). The Diabetic Retinopathy Screening Executive, on behalf of NHS Scotland, endorsed this view and recommended to the Scottish Government that automated grading should be implemented within each health board in Scotland. Computer software was developed by the private sector for use by the NHS and was piloted before national implementation.

HERU researchers involved in this research project: Paul McNamee, Graham Scotland

External Collaborators: J Olson, S Philip, G Williams (Grampian Retinal Screening Programme); P Sharp, K Goatman, A Fleming (Bio-Medical Physics and Bio-Engineering, University of Aberdeen); G Prescott (Population Health, University of Aberdeen); G Leese (Tayside Health Board); K Kyaw Swa (NHS Lothian); M Virdi (NHS Lanarkshire) and W Wykes (South Glasgow University Hospitals NHS Trust)

Publications

Fleming, A.D., Goatman, K.A., Philip, S., Williams, G.J., Prescott, G.J., Scotland, G., McNamee, P., Leese, G.P., Wykes, W.N., Sharp, P.F., Olson, J.A. and Scottish Diabetic Retinopathy Clinical Research Network (2010) 'The role of haemorrhage and exudate detection in automated grading of diabetic retinopathy', British Journal of Ophthalmology, 94(6), 706-711.

Olson, J., Philip, S., Williams, G., Sharp, P., Goatman, K., Fleming, A., McNamee, P., Prescott, G., Leese, G., Kyaw Swa, K., Virdi, M. and Wykes, W. (2010) The role of automated level 2 grading within the diabetic retinopathy screening collaborative network. Final Report to the Chief Scientist Office, Scottish Government Chief Scientist Office.

Scotland, G., McNamee, P., Fleming, A.D., Goatman, K.A., Philip, S., Prescott, G.J., Sharp, P.F., Williams, G.J., Wykes, W., Leese, G.P. and Olson, J.A. (2010) 'Costs and consequences of automated algorithms versus manual grading for the detection of referable diabetic retinopathy', British Journal of Ophthalmology, 94(6), 712-719

Presentations

Scotland, G., McNamee, P., Fleming, A. D., Goatman, K. A., Philip, S., Prescott, G. J., Sharp, P. F., Williams, G. J., Wykes, W., Leese, G. P., Olson, J. A. on behalf of the Scottish Diabetic Retinopathy Clinical Research Network (2010) 'Costs and consequences of automated algorithms versus manual grading for the detection of referable diabetic retinopathy', World Ophthalmology Congress, Berlin, June 2010.