Using Artificial Intelligence, researchers are now able to identify the back of the eye images.
Scientists have utilized Artificial intelligence (AI) to produce a far more accurate and in-depth method for analyzing images of the rear of this eye, a prior which may help ophthalmologists better identify and monitor eye diseases as glaucoma, and age-related macular degeneration.
In the study, released in the Scientific journal report, the scientists looked for a new way of analyzing images from a state-of-the-art instrument known as the Optical Coherence Tomography (OCT). The scientists, together with those from the Queensland Faculty of Technology (QUT) found Australia, explored a range of machine learning strategies to analyze OCT pictures.
The retina and the choroid are the two main tissue layers at the back of the eye and researchers tried extracting images from these two layers. Optometrists and ophthalmologists commonly use OCT, takes high-resolution images of the eye which is also cross-sectional, showing different tissue layers.
These pictures, the study observed, are actually of tissues about 4 microns thick. To put that for perspective, the human hairs are 100 microns heavy, the scientists said. OCT may be utilized to map and monitor the thickness of these tissue levels in the eye, assisting clinicians to identify eye diseases, stated David Alonso Caneiro, lead author of the analysis from QUT.
“The area between the retina and the sclera is called the choroid, that contains the major blood vessel which provides oxygen and nutrients to the eye,” The standard imaging processing methods used with OCT, he added, identified and analyzed the retinal cells levels properly, but couple of clinical OCT instruments had the software program that analyzed choroidal cells.
“So a deep learning network was made to learn the features of the image and then to automatically and accurately define the boundaries of the retina and the choroid.” He said.
In an 18 month study of 101 children with healthy eyes, researchers collected OCT chorioretinal eye scans. Using these images, researchers trained the software to detect patterns and then to define the choroid boundaries.
“The machine learning programme is much more reliable and accurate when compared to the effects with typical image analysis techniques”, said Alonso-Caneiro. Being in a position to analyze OCT images has enhanced our understanding of eye tissue changes related to normal eye growth, aging, refractive mistakes, and eye disease”.
He added that having more efficient information from these pictures of the choroid was clinically crucial and for understanding much more about the eye through research. Based on Alonso Caneiro, the new method might offer a means to higher map and monitor changes in the choroid cells, and possibly diagnose eye diseases earlier. He included that the brand new programme was discussed with eye researchers within Australia and overseas, and was optimistic that commercial OCT instruments could incorporate it.
Complex Eye Scans now easier using AI.
Analytics Jobs