Retinal Image Analysis
Retinal imaging has rapidly grown within ophthalmology in the past twenty years. The availability of cheap cameras to take direct images of the retina, fundus photography, makes it possible to examine the eye for the presence of many different eye diseases with a simple, non-invasive method. Within DIAG, research has focused on automatic early detection of diabetic retinopathy from fundus photographs. We are currently also applying the computer-aided detection and quantification techniques we have developed to diagnosis and quantification of macular degeneration.
Diabetic Retinopathy (DR) is the major cause of blindness and vision loss in developed countries among the working population. This ocular disease is a disorder of the retinal vasculature that eventually develops in nearly all patients with diabetes. There exist treatment that can stop the progression of the disease and therefore timely detection is crucial to avoid irreversible eyesight problems.
|Left: Scene viewed by a person with normal vision. Right: scene viewed by a person with diabetic retinopathy (source).
To ensure timely treatment, effective screening programs require the examination of the retina of diabetic patients at least once a year. The current practice of DR screening is based on manual examination of fundus photographs by human experts. It requires trained ophthalmologists to examine the retinal images, searching for retinal lesions.
Screening programs are costly and time consuming due to the high prevalence of diabetes and the shortage of specialists. In the Netherlands approximately 500,000 persons are affected by diabetes, and this number is expected to increase to over 700,000 by 2030. They need to undergo retinal examinations resulting in a huge amount of images that need to be reviewed. This puts an enormous burden on ophthalmologists and increases waiting lists, and may compromise the quality of health care.
Age-related macular degeneration (AMD) is the leading cause of irreversible vision loss among the elderly in developed countries and the third leading cause worldwide. AMD is a disease associated with aging that gradually destroys sharp, central vision. Because older people represent an increasingly larger percentage of the general population, vision loss from AMD is a growing public health problem. In view of the aging population, the number of patients with AMD is expected to double within the next 15 years.
|Left: Scene viewed by a person with normal vision. Right: scene viewed by a person with age-related macular degeneration (source).
At early stages of AMD visual symptoms are usually unnoticeable, while severe loss of central vision is prevalent at later stages. Fortunately, treatment possibilities for AMD have vastly improved the last decade. To prevent the progression of the disease, timely detection is crucial.
Fundus photography is an essential part of the current protocol for the diagnosis of DR and AMD. To obtain a correct diagnosis and grading of the diseases, several modalities are used in clinical practice, such as color photography, fluorescein angiography, fundus autofluorescence or optical coherence tomography.
|Fundus photography. Left: Fundus photography without pathology. Right: Fundus photography containing lesions related to Diabetic Retinopathy.
|Fundus photography. Left: Color image from a patient with AMD. Right: Fluorescein angiography from the same patient. The transparent square on the color images represents the corresponding area cover by the fluorescein angiography.
The current standard for grading fundus photographs involve the manual estimation of lesion loads and locations, which is time-consuming and has relatively low intra- and interobserver reproducibility.
Computer-aided detection for the diagnosis of eye diseases
A CAD system for DR screening is a realistic option to lower the workload of ophthalmologists and reduce the costs of health care. We have developed a CAD system for automatic DR screening which allows the identification of diabetic patients that need to be referred to an ophthalmologist. The system takes into account normal anatomy and abnormalities of multiple types as well as whether a reliable analysis can be produced depending on the quality of the image.
A CAD system for AMD diagnosis and grading is a valuable tool in avoiding the shortcomings of manual grading and reduce the ophthalmologists' workload. One of the goals of our project is to develop a CAD system able to diagnose and grade AMD using the different image modalities available in clinical practice.
|Developed retinal viewer for visualization and quantification of retinal lesions.
- A cost-effective solution for the prevention of blindness using Computer-Aided Diagnosis and fundus photography (PI Clarisa Sanchez; budget €280,000; period June 2011 - June 2015). Funded by NWO and ZonMW.
- Towards Intelligent Machines: Design of Dynamic Computer-Aided Diagnosis Systems (PI Bram van Ginneken; budget €600,000; period May 2008 - May 2013). Funded by NWO Exact Sciences, Radboud University Nijmegen Medical Centre, University Medical Center Utrecht.
- IOP Image Processing. Period 2003 - 2006. Funded by Ministry of Economic Affairs, The Netherlands.
- C.I. Sánchez, M. Niemeijer, I. Isgum, A.V. Dumitrescu, M.S.A. Suttorp-Schulten, M.D. Abràmoff and B. van Ginneken. "Contextual computer-aided detection: Improving bright lesion detection in retinal images and coronary calcification identification in CT scans", Medical Image Analysis 2012;16:50-62. Abstract/PDF DOI PMID Citations found: 3
- M.J.J.P. van Grinsven, B. van Ginneken and C.I. Sánchez. "Web-based workstation for the analysis of color fundus images", in: ISBI Medical Image Analysis Workshop, 2012. Abstract
- M.J.J.P. van Grinsven, J.P.H. van de Ven, Y.T.E. Lechanteur, B. van Ginneken, C.B. Hoyng, T. Theelen and C.I. Sánchez. "Automatic Drusen Detection and Quantification for Diagnosis of Age-Related Macular Degeneration", in: ARVO-ISIE, 2012. Abstract
- C.I. Sánchez, M. Niemeijer, A.V. Dumitrescu, M.S.A. Suttorp-Schulten, M.D. Abràmoff and B. van Ginneken. "Evaluation of a Computer-Aided Diagnosis system for Diabetic Retinopathy screening on public data", Investigative Ophthalmology and Visual Science 2011;52:4866-4871. Abstract/PDF DOI PMID Citations found: 5
- M. Niemeijer, B. van Ginneken, M.J. Cree, A. Mizutani, G. Quellec, C.I. Sánchez, B. Zhang, R. Hornero, M. Lamard, C. Muramatsu, X. Wu, G. Cazuguel, J. You, A. Mayo, Q. Li, Y. Hatanaka, B. Cochener, C. Roux, F. Karray, M. Garcia, H. Fujita and M.D. Abràmoff. "Retinopathy online challenge: automatic detection of microaneurysms in digital color fundus photographs", IEEE Transactions on Medical Imaging 2010;29:185-195. Abstract/PDF DOI PMID Citations found: 77
- C.I. Sánchez, M. Niemeijer, M.D. Abràmoff and B. van Ginneken. "Active learning for an efficient training strategy of computer-aided diagnosis systems: application to diabetic retinopathy screening", in: Medical Image Computing and Computer-Assisted Intervention, volume 6363 of Lecture Notes in Computer Science, 2010, pages 603-610. Abstract/PDF DOI PMID Citations found: 4
- C.I. Sánchez, M. Niemeijer, M.S.A. Suttorp-Schulten, M.D. Abràmoff and B. van Ginneken. "Improving hard exudate detection in retinal images through a combination of local and contextual information", in: IEEE International Symposium on Biomedical Imaging, 2010, pages 5-8. Abstract/PDF DOI Citations found: 8
- M. Niemeijer, M.D. Abràmoff and B. van Ginneken. "Fast detection of the optic disc and fovea in color fundus photographs", Medical Image Analysis 2009;13:859-870. Abstract/PDF DOI PMID Citations found: 32
- M. Niemeijer, M.D. Abràmoff and B. van Ginneken. "Information fusion for diabetic retinopathy CAD in digital color fundus photographs", IEEE Transactions on Medical Imaging 2009;28:775-785. Abstract/PDF DOI PMID Citations found: 34
- C.I. Sánchez, M. Niemeijer, T. Kockelkorn, M.D. Abràmoff and B. van Ginneken. "Active learning approach for detection of hard exudates, cotton wool spots, and drusen in retinal images", in: Medical Imaging, volume 7260 of Proceedings of the SPIE, 2009, pages 72601I1-72601I8. Abstract/PDF DOI Citations found: 3
- M.D. Abràmoff, M. Niemeijer, M.S.A. Suttorp-Schulten, M.A. Viergever, S.R. Russell and B. van Ginneken. "Evaluation of a system for automatic detection of diabetic retinopathy from color fundus photographs in a large population of patients with diabetes", Diabetes Care 2008;31:193-198. Abstract/PDF DOI PMID Citations found: 97
- M. Niemeijer, M.D. Abràmoff and B. van Ginneken. "Segmentation of the Optic Disc, Macula and Vascular Arch in Fundus Photographs", IEEE Transactions on Medical Imaging 2007;26:116-127. Abstract/PDF DOI PMID Citations found: 94
- M. Niemeijer, B. van Ginneken, S.R. Russel, M.S.A. Suttorp-Schulten and M.D. Abràmoff. "Automated Detection and Differentiation of Drusen, Exudates, and Cotton-Wool Spots in Digital Color Fundus Photographs for Diabetic Retinopathy Diagnosis", Investigative Ophthalmology and Visual Science 2007;48:2260-2267. Abstract/PDF DOI PMID Citations found: 105
- M. Niemeijer, M.D. Abràmoff and B. van Ginneken. "Image structure clustering for image quality verification of color retina images in diabetic retinopathy screening", Medical Image Analysis 2006;10:888-898. Abstract/PDF DOI PMID Citations found: 41
- M. Niemeijer. "Automatic detection of diabetic retinopathy in digital fundus photographs", PhD thesis, Utrecht University, The Netherlands, 2006. Abstract/PDF URL
- M. Niemeijer, B. van Ginneken, J. Staal, M.S.A. Suttorp-Schulten and M.D. Abràmoff. "Automatic Detection of Red Lesions in Digital Color Fundus Photographs", IEEE Transactions on Medical Imaging 2005;24:584-592. Abstract/PDF DOI PMID Citations found: 167
- J.J. Staal, M.D. Abràmoff, M. Niemeijer, M.A. Viergever and B. van Ginneken. "Ridge Based Vessel Segmentation in Color Image of the Retina", IEEE Transactions on Medical Imaging 2004;23:501-509. Abstract/PDF DOI PMID Citations found: 688