Diabetes imposes a large public health burden and one of its many complications is diabetic retinopathy. Early detection of diabetic retinopathy is an important problem because, in spite of well-known treatment benefits, 50% of Americans with diabetes do not undergo regular eye examinations due mainly to demographic, provider, economic and educational factors. This application is in response to a Request for Applications issued by the National Eye Institute (NEI) to develop, apply, and evaluate a noninvasive technology that is practical, affordable, and accessible so that patients with diabetes can benefit from remote site disease screening. We have developed an internet-based system specifically designed for this purpose. It consists of an automated imaging system, the DigiScope, placed in the primary care environment in conjunction with a remote expert reading center. The DigiScope has been validated, in a specialty clinic, against the gold-standard for the detection of diabetic retinopathy, seven-field stereo fundus photography, for the detection of diabetic retinopathy requiring referral to an ophthalmologist. The DigiScope has been implemented in a variety of primary care physicians' offices and more than 12,000 patients who did not have an eye examination in the previous year have been screened. The initial implementation has shown that this system is economically feasible, convenient, easy to use, and increases the rate of screening. The implementation has also led to the identification of improvements that are necessary to increase the utility of the system.
The aim of the proposed research is to upgrade the DigiScope to shorten the imaging time, decrease the number of unreadable images and improve the portability of the instrument. The upgraded DigiScope will be tested for technical performance in an ophthalmic clinic. This will be followed by a comprehensive evaluation study in the primary care environment. The protocol will comply with methodologies used in multi-center studies sponsored by the NEI. The impact of DigiScope screening on the overall rate of diabetic eye examinations will be determined and the compliance with recommended follow-up will be reviewed. Finally, since a telemedicine screening system requires reliable performance of the reading center, training and certification of readers, and quality controls are essential. We will lay the concepts and specifications for a computer based training program for education, certification and quality control of readers. ? ? ?
| Moscaritolo, Michele; Jampel, Henry; Zimmer-Galler, Ingrid et al. (2010) A computer algorithm to quantitatively assess quality of digital optic disc images. Ophthalmic Surg Lasers Imaging 41:279-84 |
| Moscaritolo, Michele; Knezevich 3rd, Frederick P; Zimmer-Galler, Ingrid et al. (2010) A machine vision method for automated alignment of fundus imaging systems. Ophthalmic Surg Lasers Imaging 41:607-13 |
| Moscaritolo, Michele; Jampel, Henry; Knezevich, Frederick et al. (2009) An image based auto-focusing algorithm for digital fundus photography. IEEE Trans Med Imaging 28:1703-7 |
| Zimmer-Galler, Ingrid E; Zeimer, Ran (2009) Telemedicine in diabetic retinopathy screening. Int Ophthalmol Clin 49:75-86 |
