Our Focus

In a collaborative, multicenter study involving researchers from India and the Netherlands, Dr. Jay Sheth, Head of Retina and Clinical Research at Shantilal Shanghvi Eye Institute (SSEI), was part of a team that developed and validated an artificial intelligence (AI) model aimed at improving the diagnosis of retinal emergencies using ocular ultrasound.

What is it about?

When patients present with sudden flashes, floaters, or vision loss, quick and accurate diagnosis is critical. Two common culprits are retinal detachment (RD), a sight-threatening emergency, and posterior vitreous detachment (PVD), a more benign condition where the vitreous, the jelly-like substance inside the eye, separates from the retina.

Both conditions can appear similar on B-scan ultrasonography, showing as membranes attached to the optic disc. Differentiating between them, especially in busy or resource-limited settings, can be challenging.

To address this, the research team developed a deep learning model based on Vision Transformer (ViT) technology. Trained on hundreds of ultrasound images, the AI tool could accurately classify eyes as RD, PVD, or normal, achieving over 95% accuracy in each category, closely matching expert-level diagnosis.

Why it matters

Retinal detachment requires urgent treatment to prevent permanent vision loss. This AI tool has the potential to support clinicians in making faster and more accurate decisions, especially in high-volume or remote settings where access to retinal specialists may be limited.

By combining clinical insight with technological innovation, this study highlights how AI can enhance early detection and timely intervention. SSEI is proud to be part of this global effort to improve outcomes for patients facing vision-threatening emergencies.

Citation

Bhatt VD, Shah N, Bhatt DC, Dabir S, Sheth J, Berendschot TTJM, Erckens RJ, Webers CAB. Development and Evaluation of a Deep Learning Algorithm to Differentiate Between Membranes Attached to the Optic Disc on Ultrasonography. Clin Ophthalmol. 2025 Mar 18;19:939-947. doi: 10.2147/OPTH.S501316.