Published in Products

Topcon and RetiSpec partner on ocular AI diagnostic tool for Alzheimer's

This is editorially independent content
4 min read

Topcon Healthcare, Inc. and RetiSpec, Inc. are collaborating to bring RetiSpec's eye diagnostic artificial intelligence (AI) technology to the commercial market.

First up: these companies.

As a division of the Japanese optical equipment manufacturer Topcon Corporation, Topcon Healthcare is a global provider of medical devices and software solutions to the ophthalmic community.

In recent news: The company announced its new CEO and president: Ali Tafreshi (details here).

Toronto, Canada-based RetiSpec is a medical imaging company that uses advanced AI to detect neurodegenerative diseases (e.g., Alzheimer’s disease [AD]) via hyperspectral imaging of the eye.

Note: The start-up was spun out of the University of Minnesota in 2018.

Its key technology: a proprietary AI-based system that photographs a patient’s retina via a hyperspectral camera.

Why hyperspectral?

This is based on technology developed by researchers at the University of Minnesota’s Center for Drug Design, which included the underlying method of using retinal hyperspectral imaging to detect the pathology of AD via an eye exam.

Fast forward to 2014: Investigators developed a novel tissue spectroscopy method to test the effectiveness of a potential AD therapeutic and assessed the potential capabilities of hyperspectral imaging for detecting AD pathology.

And the result of this?

Based on evidence from “cell line studies, ex-vivo rodent brain studies, and ex-vivo human studies,” toxic soluble amyloid-β (Aβ) aggregates—a core biomarker of AD—were detected using hyperspectral imaging.

Note: Transient soluble oligomers of Aβ (a protein) are known to abnormally accumulate prior to insoluble plaque formation and cognitive impairment in AD.

How was this validated?

The investigators clinically validated this technology by predicting brain Aβ status in patients with and without AD in multiple prospective studies (single and multi-site designs), according to the company.

Which leads us to RetiSpec’s AI solution …

The basis for it! With this knowledge of Aβ, investigators developed a specialized hyperspectral camera attached to standard fundus imaging equipment.

The solution incorporates existing retinal imaging cameras already available in most clinics (so the company says).

The intention: to provide a non-invasive, accurate, and scalable solution for detecting early AD.

And what kind of images are produced?

Per RetiSpec: data-rich images of the retina via high-resolution retinal imaging, with each pixel containing over 100 colors (in lieu of the standard red, blue, and green in regular color images).

How is AI incorporated?

The system uses AI algorithms trained on other retinal scans from AD patients to analyze the output of images.

Now to this partnership … how is Topcon’s tech being integrated?

First: Take note that RetiSpec’s AI-based technology is currently only for research purposes.

Once commercially available, however, the companies plan to integrate the device with Topcon Harmony, a clinical image and data management solution designed to streamline clinical practice workflow and organize all ophthalmic clinical data in one screen.

The expectation: to enable early AD detection and more timely medical management and access to treatment, according to the companies.

So when might RetiSpec’s tech be commercialized?

While there’s no specific timeframe as of yet, Topcon noted that this collaboration is expected to accelerate the commercialization and scale of the system.

How would you rate the quality of this content?