Web-based grading of compressed stereoscopic digital photography versus standard slide film photography for the diagnosis of diabetic retinopathy.

PURPOSE: To evaluate the ability of Joint Photographic Experts Group (JPEG) compressed, stereoscopic, digital photography to identify clinical levels of diabetic retinopathy, detect clinically significant macular edema (CSME), and make appropriate referral recommendations as compared with Early Treatment Diabetic Retinopathy Study (ETDRS) standard film photography. DESIGN: Prospective, clinic-based, comparative photographic format validation study. PARTICIPANTS: Two hundred four eyes from 102 consecutive, new, diabetic patients with a median duration of diabetes of 12.5 years were enrolled and analyzed. METHODS: After pupillary dilation, a trained ophthalmic photographer obtained 2 sets of images: standard ETDRS, stereoscopic 7-field 35-mm film photographs and high-resolution, digital images of the same 7 standard fields (stereoscopic images of fields 1 and 2). Two masked readers graded both sets of images. The 35-mm film photographs were graded by hand using a written template. Digital photographs were compressed, uploaded to a web site, and graded by a web-based, computer-assisted ETDRS algorithm. The ETDRS level of diabetic retinopathy, presence of retinal thickening, and referral recommendation based on these 2 diagnoses were recorded and compared. MAIN OUTCOME MEASURES: The presence of CSME, ETDRS level of diabetic retinopathy, and referral threshold diabetic retinopathy. RESULTS: Film and compressed digital grading levels were compared using a simplified ETDRS categorization scheme. Film and digital gradings were highly correlated with exact agreements for level of diabetic retinopathy, CSME, and referral thresholds >87% and kappa levels >0.71. McNemar's testing found no statistically significant difference between compressed digital images and film when comparing referral thresholds (defined as the presence of CSME and/or ETDRS level > or = 61; P = 0.76). CONCLUSIONS: A modified ETDRS protocol (stereoscopic images of fields 1 and 2 only) with 16:1 JPEG image compression and computer-assisted ETDRS grading algorithm has excellent reproducibility when compared to standard ETDRS stereoscopic slide-film photography. An internet-based teleophthalmology system can correctly and reliably (kappa = 0.78) identify patients requiring referral for CSME and proliferative diabetic retinopathy (PDR; ETDRS level > or = 61).

Benefits of stereopsis when identifying clinically significant macular edema via teleophthalmology.

BACKGROUND: The need to incorporate stereopsis into a teleophthalmology system is controversial. Detection of hard exudate in the macula is suggested by some as an adequate surrogate for direct observation of retinal thickening. This study was designed to determine how accurate the detection of hard exudate is as a surrogate for stereoscopic detection of clinically significant macular edema (CSME). METHODS: 120 patients with diabetes underwent clinical retinal examination with contact-lens biomicroscopy by a retinal specialist. The presence or absence of CSME was recorded. On the same day as clinical grading, 30 degrees stereoscopic digital photographs of the macula were captured. At least 2 months after clinical examination, the digital images were viewed by masked graders for the presence or absence of hard exudate and retinal thickening. RESULTS: 207 eyes of 106 patients had complete data sets for both diagnostic modalities. The sensitivity of hard exudate (93.9%) in predicting the presence of CSME was similar to that of direct stereoscopic observation of retinal thickening (90.9%), with p = 0.5. On the other hand, digital stereopsis was significantly more specific (92.9%) than was hard exudate (81.6%) in predicting the presence of CSME (p < 0.001). This difference was maintained even when controlling for image quality. INTERPRETATION: Although the presence of hard exudate within the macula is a sensitive surrogate marker for CSME, it is less specific than stereoscopic evaluation. Any American Telemedicine Association category 3 teleophthalmology system that utilizes hard exudate as a surrogate marker for CSME may refer patients unnecessarily for clinical evaluation.

JPEG compression of stereoscopic digital images for the diagnosis of diabetic retinopathy via teleophthalmology.

BACKGROUND: Canada's vast size and remote rural communities represent a significant hurdle for successful monitoring and evaluation of diabetic retinopathy. Teleophthalmology may provide a solution to overcome this problem. We investigated the application of Joint Photographic Experts Group (PEG) compression to digital retinal images to determine whether JPEG compression could reduce file sizes while maintaining sufficient quality and detail to accurately diagnose diabetic retinopathy. METHODS: All 20 patients with type 2 diabetes mellitus assessed at a 1-day teleophthalmology clinic in northern Alberta were enrolled in the study. Following pupil dilation, seven 30 degrees fields of each fundus were digitally photographed at a resolution of 2008 x 3040 pixels and saved in uncompressed tagged image file format (TIFF). The files were compressed approximately 55x and 113x their original size using JPEG compression. A reviewer in Edmonton randomly viewed all original TIFF images along with the compressed JPEG images in a masked fashion for image quality and for specific diabetic retinal pathology in accordance with Early Treatment Diabetic Retinopathy Study standards. The level of diabetic retinopathy and recommendations for clinical follow-up were also recorded. Exact agreement and weighted kappa statistics, a measure of reproducibility, were calculated. RESULTS: Exact agreement between the compressed JPEG images and the TIFF images was high (75% to 100%) for all measured variables at both compression levels. Reproducibility was good to excellent at both compression levels for the identification of diabetic retinal abnormalities (K = 0.45-1), diagnosis of level of retinopathy (kappa = 0.73-1) and recommended follow-up (kappa = 0.64-1). INTERPRETATION: The application of JPEG compression at ratios of 55:1 and 113:1 did not significantly interfere with the identification of specific diabetic retinal pathology, diagnosis of level of retinopathy or recommended follow-up. These results indicate that JPEG compression at ratios as high as 113:1 has the potential to reduce storage requirements without interfering with the accurate and reproducible teleophthalmologic diagnosis of diabetic retinopathy. This pilot project demonstrates the potential for JPEG compression within a digital teleophthalmology viewing system.