Michael D. Abràmoff, MD, PhD

Michael D. Abràmoff, MD, PhD, is an ophthalmologist in the Department of Ophthalmology and Visual Sciences. He received his M.D. and medical retina fellowship in Amsterdam, the Netherlands, and his ophthalmology residency and PhD in image analysis in Utrecht, also in  the Netherlands. He studies image analysis to prevent blindness and visual loss from retinal disease, focusing on the ‘Big Three of Blindness’: Diabetic Retinopathy, Age Related Macular Degeneration and glaucoma. The retina is the tissue in the back of the eye that can be imaged with retinal color cameras, optical coherence tomography and other techniques.

Using image analysis, he puts numbers on these retinal images. This allows the image to be treated in the same objective and reproducible manner as any other clinical measure, such as weight, blood pressure, or refractive power. Thus, retinal images can be used for early diagnosis, improved management, discovery of new mechanisms of disease, and phenotype discovery. He is very excited to use the newest findings in brain research of vision to improve his image analysis algorithms.

From its very nature, this type of imaging research is extremely collaborative, and he has been fortunate enough to be in a position to bridge the gap between leading clinicians and engineers here in Iowa: Professor L.H. Alward (Ophthalmology and Visual Sciences), Professor R.K. Kardon (Ophthalmology and Visual Sciences), Associate Professor Y.H. Kwon (Ophthalmology and Visual Sciences), Professor R.L. Reinhardt (Biomedical Engineering), Professor S.R. Russell (Ophthalmology and Visual Sciences), Professor E.M. Stone (Ophthalmology and Visual Sciences), Professor T.E. Scheetz (Bioinformatics and Ophthalmology and Visual Sciences) and Professor M.S. Sonka (Electrical and Computer Engineering, and Ophthalmology and Visual Sciences), visiting Professor B. van Ginneken (Electrical and Computer Engineering) and Professor X. Wu (Electrical and Computer Engineering and Department of Radiotherapy) .

Funded by the National Institutes of Health, he is developing, with his collaborators, automated detection of diabetic retinopathy, a feared complication of diabetes, and the second most important cause of blindness. This system uses retinal images to assist eye care providers with the early detection of diabetic retinopathy in the 18 million Americans with diabetes. He has shown that the system can detect hemorrhages and other damage to the retina caused by diabetes at a level comparable to that of ophthalmologists. Recently, the system was tested on 40,000 retinal images from patients with diabetes, and was compared to the diagnosis of ophthalmologists, with favorable results. Currently the algorithms are being improved to detect an even larger variety of lesions.

 Figure: Example of fully automated detection of lesions and retinal vessels
in a patient with diabetic retinopathy. System gives the probability of each lesion being a hemorrhage, vessel or exudates, as well as each patient’s images containing diabetic retinopathy. Therefore, the system can have different disease detection sensitivities.

Funded by the Veterans’ Administration, he is developing a similar system for the automated detection of glaucoma, the third most common cause of blindness. He has recently shown that this system, can segment the optic nerve at the back of the eye at a performance level comparable to ophthalmologists.

Funded by the National Institutes of Health, he and his team are also developing a low cost camera for imaging the retina, which is more patient friendly, and easier to use by rural nurses and physicians. Note: Dr. Abramoff has a share in the company, iOptics LLC, commercializing a marketable version of this camera.

He and his coworkers have also developed a new approach for fully three-dimensional analysis of retinal tomographic images – spectral domain OCT (Optical Coherence Tomography). Three-dimensional OCT is a technique that became available in the summer of 2007, which can image the issues within the retina. He is currently further improving these 3D OCT analysis algorithms to guide treatment for Age Related Macular Degeneration and Diabetic Retinopathy.

Finally, he and his colleagues have developed a novel approach for retinal phenotype discovery that is unbiased and objective.