RetInSight, a leader in AI-assisted retinal biomarker image analysis, has received its EU Quality Management System certificate according MDR 2017/745 on December 13, 2021.(more…)
Heidelberg, Germany – Heidelberg Engineering GmbH and RetInSight GmbH intend to interface the RetInSight AI-based fluid monitor application with the Heidelberg Engineering product portfolio, using cloud exchange and application marketplace technologies.(more…)
Regulatory approval of a novel artificial intelligence-based tool that provides precise measurement of retinal fluid may introduce major paradigm changes in the monitoring of patients with neovascular age-related macular degeneration.
“It is a precision management tool, a VEGF meter, which is what Philip Rosenfeld was asking for when anti-VEGF therapies started. More precisely, it is a fluid meter and is now ready for use in clinical practice,” Ursula Schmidt-Erfurth, MD, PhD, told Healio/OSN.
As a tool that every eye specialist can introduce in their daily routine practice, the RetInSight fluid monitor is a milestone in a long-standing research pathway of development and validation of deep learning methods to support the diagnosis and management of retinal disease. Schmidt-Erfurth and her group at the University of Vienna were pioneers in this field and worked on this project for almost a decade.
“Anti-VEGF therapy has been in our hands since 2006. Nevertheless, the real-world outcomes are far below what we expected from the clinical studies, and severe vision loss and blindness have increased by 24% during the last 10 years. This is a learning and a warning,” she said.
In addition, there are no shared evidence-based criteria for re-treatment and injection frequency, and there is no consensus among retina physicians on the best treatment regimen to reduce patient burden, promote compliance and ensure treatment efficacy. The current parameters to determine the treatment interval are not sufficiently objective or reliable, according to Schmidt-Erfurth.
“We have no standard. Everybody is doing their best with a huge variability. There is no real criterion for even highly educated investigators to say whether a patient should be treated or should be put forward to a longer interval. Treatment decisions are based on central retinal thickness, which we know does not correlate with fluid volumes,” Schmidt-Erfurth said.
The RetInSight fluid monitor is a “fluid activity meter,” a detection algorithm that objectively identifies, localizes and quantifies fluid in OCT images on a three-dimensional pixel basis.
“So, we have accurate volumes in nanoliters, we have accurate locations within the central macular area, and each OCT pixel is labeled according to its fluid type as intraretinal fluid (IRF), subretinal fluid (SRF) and/or pigment epithelium detachment when present,” Schmidt-Erfurth said. “It is a personalized, automated, real-time system to identify disease activity and the response to therapy.”
Central retinal thickness is currently used as a surrogate OCT biomarker for disease activity, but particularly in neovascular AMD, what the instrument measures is not the actual 3D fluid pooling. There is no assessment of tridimensionality and no perception of quantity. And yet, accurate, dynamic fluid representation and fluid control are critical in neovascular AMD. A study by Chakravarthy and colleagues showed worsening of visual acuity over 2 years associated with dynamic volume fluctuations in retinal fluid volumes during the maintenance phase.
“It’s the ups and downs that cause the vision loss, and these changes need to be objectively measured. They cannot be assessed by even the best expert,” Schmidt-Erfurth said. “With the fluid monitor, we are introducing a standard, and that means that we have objective criteria that allow us to treat as much as needed but as little as possible and to do that at the right time interval.”
Schmidt-Erfurth does not agree with a large part of the community that has “fled toward treat-and extend,” which means treating a dry retina in many cases. In medicine, giving a drug when it is not needed is never a good idea, she said, and a drug that has to be injected into people’s eyes hundreds of times over a lifetime should preferably be administered when there is active fluid.
“And fluid has to be measured, so that the new standard would be not to treat a dry retina. The other standard would be to learn that fluid comes in different compartments. Even this very basic truth is not respected in the standards we have now. We just treat any fluid, although we are learning that IRF and SRF have distinct functional implications and that IRF is more consistently associated with worse visual outcomes. We may introduce a standard that only treats dynamic fluid and not persistent SRF, which may even be good as long as it remains stable,” Schmidt-Erfurth said.
The new tool is meant to be a practical support for clinicians in their treatment decision-making process. It is a cloud-based software compatible with all leading OCT systems, starting with the Heidelberg Engineering Spectralis.
“Over the follow-up of your patients, you can at any time see a curve that shows exactly the biological activity and treatment response, and you no longer have to search and compare with previous images. It is very much hands-on and makes our lives as retinal experts or general ophthalmologists much easier,” Schmidt-Erfurth said. “And it brings the patients on board who are fully informed about their measurements and will increase compliance.”
It is also an incentive to the adoption of shared-care practices, resulting in a more efficient workflow and more time spent with the patient.
“The fluid monitor will bring clinical trial quality to routine clinical care, and at last we should be able to bridge the gap between clinical trial results and real-world outcomes,” Schmidt-Erfurth said.
In addition, cloud-based fluid analysis has the potential to introduce important changes in clinical trials, drastically reducing the variability involved with data collection and analysis. Accessibility of real-time disease condition will speed up recruitment and patient screening, allowing immediate exchange of outcome data with optimal protocol adherence during treatment and a completed data closing right at the end of the study.
“Sponsors, reading centers and everybody involved can immediately get the response of every single patient, and at the end of the study, all the data will be timely available, reliable and transparent. It will save a lot of manpower and budget also in clinical studies,” Schmidt-Erfurth said.
The RetInSight fluid monitor has received the CE mark and Medical Devices Regulation label, currently the highest regulatory approval in Europe. It has been launched there, and an early access program is in place for key opinion leaders.
“We are interested in getting the comments of experts who will be using the fluid monitor in their own practice, and there will be an official launch at Euretina,” Schmidt-Erfurth said.
At EURETINA and ESCRS 2022, Heidelberg Engineering will present a new feature of its growing healthcare IT portfolio. Heidelberg AppWay, a novel gateway solution for secure data exchange, will allow seamless access to ophthalmic ‘apps’ that aid clinical decision making. For the launch of this workflow solution, Heidelberg Engineering has partnered with the companies RetInSight and RetinAI to provide eye care specialists with innovative AI software systems that support retinal diagnostics and therapeutic monitoring.
Heidelberg Engineering is expanding its healthcare IT portfolio with the introduction of Heidelberg AppWay. This latest addition to the company’s ecosystem offers secure access to ophthalmic software analytics and AI-powered support systems. It will be embedded into the image management platform HEYEX 2, providing an innovative workflow solution that facilitates seamless exchange of ophthalmic image data with AI-driven analysis services to deliver additional clinical insights as well as treatment monitoring tools for enhanced patient care.
“With Heidelberg AppWay, we aim to support clinicians seeking AI-driven solutions to aid their decision making, and researchers interested in innovative analysis tools,” says Krysten Williams, Global Head of Marketing and Education at Heidelberg Engineering. “The services of our app partners in combination with the powerful data of our imaging platforms will empower eye care professionals to uncover even more details of the eye.”
For the launch of the novel gateway solution, Heidelberg Engineering has partnered with AI software companies to provide innovative ophthalmic apps. With Heidelberg AppWay, eye care professionals can access the RetInSight Fluid Monitor, for example. From any practice or clinic, clinicians can send routine images from their SPECTRALIS platform to the cloud service for a detailed, AI-powered analysis of disease activity and therapeutic benefit. While the patient is still on site, they receive a concise Fluid Monitor report indicating fluid type, distribution and quantity – in a completely secure, encrypted environment.
“This seamless workflow provides clinicians with easy access to fluid parameters – the most important biomarkers in the routine monitoring of patients with nAMD,” says Professor Ursula Schmidt-Erfurth, Chair of the Department of Ophthalmology and Optometry at the Medical University of Vienna, Austria. She is convinced that the RetInSight Fluid Monitor in combination with Heidelberg Engineering’s imaging platforms will improve the workflows of busy eye care specialists across Europe. “Availability of a fast, precise, and reliable tool will strongly improve quality and workflow in the most challenging therapeutic task in modern ophthalmology, which is the management of neovascular AMD. It will bring our ability to understand and treat disease to a substantially higher level by introducing precision into decision-making.”
AI companies on the show floor at EURETINA and ESCRS
RetInSight was founded in April 2020 as a spin-off from the Department of Ophthalmology & Optometry of the Medical University of Vienna, Austria, a global pioneer in AI-based retinal image analysis and provider of reading center services for more than 15 years.
RetInSight develops transformational AI solutions according to the highest standards for the most predominant retinal diseases of our time. The RetInSight Fluid Monitor is the first fluid activity meter for the monitoring of disease activity and therapeutic response in nAMD with CE certification under the new Medical Device Regulation. By precisely identifying, localizing and quantifying IRF, SRF and PED in the central 1-mm and 6-mm macular areas, Fluid Monitor supports the optimization and personalization of therapy, with the aim to improve patient outcomes and clinical efficiency. The AI-based clinical decision support system produces a concise result by mouse click, while the patient is on site.
With Heidelberg AppWay, Heidelberg Engineering aims to provide eye care professionals with a secure, simple, and seamless way to a growing portfolio of ophthalmic apps and AI analytics. The company will continue to add new app partners to the workflow solution.