AI Boosts Retinal Image Processing Speed by 100x

In a groundbreaking advancement reported on April 10, 2024, by the National Eye Institute and published in Communications Medicine, scientists have made a significant leap in the field of retinal imaging thanks to artificial intelligence (AI). This new method is not only 100 times faster than traditional techniques but also improves image contrast by an impressive 3.5-fold. Such progress is poised to revolutionize how researchers study and understand retinal diseases, including age-related macular degeneration (AMD).

Retinal imaging is essential for diagnosing and tracking the progression of eye diseases. Traditionally, capturing high-resolution images of cells in the eye, particularly the retinal pigment epithelium (RPE) cells, has been a lot of work. The RPE layer is crucial for eye health, supporting the retina’s light-sensing capabilities and overall function. Unfortunately, many retinal diseases occur when the RPE deteriorates. Thus, having a detailed view of this layer is vital for early detection and treatment of various conditions.

The team, led by Johnny Tam, Ph.D., at the NIH’s National Eye Institute, utilized adaptive optics (AO) combined with optical coherence tomography (OCT), a noninvasive imaging technique akin to ultrasound. While OCT is a standard tool in eye clinics, adding AO to the mix allows for a much clearer view of the retina’s cellular structure. However, this method faced challenges, primarily due to a phenomenon known as speckle, which can obscure parts of the image much like clouds can obscure views in aerial photography.

To combat this issue, Tam and his team developed a novel AI-based approach using a parallel discriminator generative adverbial network (P-GAN). This deep learning algorithm was trained with nearly 6,000 images, learning to identify and recover details hidden by speckles. Remarkably, P-GAN can produce clear, detailed images from a single capture, whereas traditional methods needed about 120 images to achieve a similar result. This not only slashes the time required for imaging but also provides images with significantly better contrast, making the fine details of the RPE layer more visible than ever before.

This AI-driven method transforms the prospects of retinal imaging, offering a faster, more detailed view and making it easier to identify early signs of diseases. It marks a shift in how AI is integrated into medical imaging, treating it as an essential part of the imaging process rather than an afterthought. As a result, AO imaging becomes more accessible for routine clinical use and research, opening new avenues for understanding and treating blinding retinal diseases.

This development represents a significant stride forward in eye care, promising quicker diagnoses and a better understanding of retinal diseases. With the potential to observe the earliest signs of conditions like AMD, this technique could lead to more effective treatments and preventative measures, ultimately saving sight for millions around the globe.

Source: Medicalxpress

---

Like this article?  Keep up to date with AI news, apps, tools and get tips and tricks on how to improve with AI.  Sign up to our Free AI Newsletter

Also, come check out our free AI training portal and community of business owners, entrepreneurs, executives and creators. Level up your business with AI ! New courses added weekly. 

You can also follow us on X