Diagnosing retinal disease, which affects over 200 million people worldwide, requires expensive and complicated analysis of the structure and function of retinal tissue. Recently, UCI developed a training algorithm which, for the first time, is able to assess tissue health from images collected using more common and less expensive optics.
• For the analysis and diagnosis of retinal tissue
• Inexpensive: The machine learning algorithm is able to extract relevant tissue information from images collected from standard and inexpensive optics.
• Small: The conditioned light source used here is small, making the platform highly portable and able to be used in remote locations.
• Powerful: Despite its small size, this platform allows for the extraction of a wide array of tissue parameters from a small set of easy-to-collect images.
Hyperspectral imaging (HSI) and dual spectral cameras are common methods used for analysis of retinal tissue, where 3D imaging and O2 levels are crucial aspects of diagnostics. However, these techniques are often performed separately and post-imaging analysis can often be long and complicated, requiring trained specialists.
Recently, UCI developed a more general platform for retinal assessment which combines the functionality of both HSI and dual spectral cameras. The platform also contains sophisticated machine learning algorithms to quickly and accurately process the multidimensional data. In this way, the method directly extracts retinal conditions such as tissue structure, O2 saturation, and health from images that otherwise would never yield this information.
Concept stage. Prototype in development.