Medical Diagnostic Software for Efficient Detection of Eye Diseases

Quickly Calculates Eye Blood-Flow Parameters Using Non-Invasive Retinal Imagery Captured During Routine Exams

This medical diagnostic software detects diseases and age-related eye deterioration by calculating eye blood flow parameters from retinal fundus images. Disease- and age-related vision impairment currently affects 4.2 million Americans over the age of 40 and may affect 8.96 million Americans by 2050. Current eye care standards require routine eye examinations that capture retinal fundus images. Medical experts must screen the images for age deterioration and vision-impairing diseases. However, analyzing retinal fundus images is time-intensive, requires specialized training, and can be prone to human error. Automated assessments of retinal fundus images captured during routine eye examinations can increase exam processing accuracy and throughput and meet growing demands for eye examinations.

Researchers at the University of Florida have developed a medical diagnostic software that detects eye diseases by calculating eye blood flow parameters from retinal fundus images that are captured during routine eye examinations.

Application

Medical diagnostic software that assists in diagnosing eye diseases by automatically screening retinal fundus images commonly collected during routine eye exams

Advantages

• Assesses retinal images more quickly than human experts, increasing the number of exams that can be processed
• Utilizes images already taken during routine, non-invasive eye exams, making the software a convenient disease detection method
• Detects disease by quantitatively evaluating eye blood flow parameters, eliminating potential bias associated with non-quantitative detection methods
• Produces comparable parameters calculated from routinely collected images of the same patient, allowing disease progression and age deterioration to be quantified objectively

Technology

This medical diagnostic software assists in diagnosing eye diseases by automatically assessing commonly collected retinal fundus images. After an image is captured during an eye exam, it is uploaded to the software through a convenient interface. The software then detects diseases by isolating the retinal vascular network from the background image of the retina and using constructal analysis to quantify blood flow parameters indicative of multiple eye diseases.