The Silicon Review
We are just two decades into the 21st century, and it is quite evident that one of the major transformative technologies and enablers for our society of this century is Artificial Intelligence (AI). The idea is clearly established that AI and associated platforms and services will completely transform the working pattern, global productivity, and lifestyle, creating enormous wealth. Conventional technology and business sectors are not the only fields that will be affected by AI. Healthcare is a field that is expected to be highly suitable for the application of AI techniques and tools. AI will enhance the quality of intelligent-decision making and automation. in our digital world, electronically-stored medical imaging data is enormous, and Deep Learning algorithms can be fed with this kind of dataset to detect anomalies and patterns.
Various firms are innovating with the AI in the healthcare sector, but one that stands out from the rest is EyRIS. The company develops and deploys AI Deep Learning System (DLS) in the healthcare industry. DLS, a breakthrough machine learning technology, utilizes representation-learning methods with multiple representation levels to process natural data in their raw form, recognizing the intricate structures in high-dimensional data. SELENA+, EyRIS’s first DLS solution, performs automated image analysis for eye diseases such as diabetic retinopathy with add-on modules for glaucoma and age-related macular degeneration. In the later part of the year, SELENA+ is slated to be adopted nationally as part of Singapore’s National Artificial Intelligence Strategy to prevent diabetic blindness. EyRIS was founded in the year 2018, and it is based in Singapore.
In conversation with Lai Teik Kin, Founding CEO of EyRIS
Q. What are the challenges you had to face while developing your solutions? How did you overcome them?
Two of the initial challenges were validating SELENA+’s algorithm and securing the data needed to do so. The acquisition of well-labeled data was restrictive due to personal data protection acts. Consequently, the research team had to brainstorm on collaboration approaches and finally navigate legal uncertainty to operate on multiple data-sharing agreements. Once we had the dataset, we also had to perform a series of data cleansing activities such as re-grading by skilled graders to ensure standardization of assessment methods such as diabetic retinopathy grading to the International Classification of DR Scale (ICDRSS), on which is SELENA+ was developed. A major milestone for SELENA+ was the JAMA paper publication, which concluded that the DLS had high sensitivity and specificity for identifying diabetic retinopathy and related eye diseases. The subsequent hurdle is regulatory approval, as SELENA+ is classified as Software as Medical Device (SaMD). Only as recent as 2017, the US FDA released its Digital Health Innovation Action Plan to offer clarity about the agency’s role in advancing safe and effective digital health technologies with other regulatory agencies around the world following suit. In its formative years, the EyRIS team has had to grapple with compliance uncertainties, territorial differences, and barriers to adoption.
In Oct 2019, the Health Science Authority of Singapore (HSA) gave us the official approval for SELENA+, which is an effective assistive tool in the diagnosis of 3 eye diseases, diabetic retinopathy (DR), Glaucoma, and Age-Related Macular Degeneration (AMD). Subsequently, we received CE Marking, which allows SELENA+ to be sold within the European Economic Area (EEAO).
Q. Eye diseases are detected and diagnosed by ophthalmologist and trained technicians. How does your solution align with the existing medical practice?
Globally, there are well over 200,000 ophthalmologists but the spread of these specialists is rather uneven. Ophthalmologists and trained eye care professionals are concentrated in more advanced counties and mainly in urban areas. In more mature economies like the UK, there are about 50 ophthalmologists per million people. In contrast, countries like Zambia and many others in Sub-Saharan Africa can only depend on one ophthalmologist per million. Studies have shown an inverse correlation between the prevalence of blindness and ophthalmologist density. The trend of poor distribution is unlikely to be reversed, and national healthcare systems are unable to continue limitless spending to keep up with ever-increasing demands. The advent of Deep Learning (DL) technology has allowed us to develop algorithms to detect diseases such as cancer and, in our case, retinal diseases accurately and in their early stages. This has provided us with a possible paradigm shift in how we can overcome financial and skilled-manpower challenges hindering the deployment of a more comprehensive screening for the population. SELENA+ can be deployed to automatically grade fundus images and filter out normal cases and at the same time, highlighting more serious cases for further attention. I envisage that AI adoption will eventually become an integral part of healthcare eco-systems.
Q. Tell in brief about the new revenue streams generated by your solutions.
As SELENA+ is a cloud-based system, the revenue stream is transaction-based – that is based on the number of test submissions.
Q. Tell us about your AI Deep Learning System and the challenges addressed by your solution.
EyRIS is uniquely positioned in 2 attractive growth industries – healthcare and technology. Artificial intelligence has made leaps and bounds in recent years and has now reached a point where the tech can actually deliver and is being adopted in industries such as banking and insurance. As AI technology was refined further to analyze and identify imaging patterns faster and more precisely, the logical application in healthcare moved in the direction of early detection and diagnosis. Conversely, diabetes is one of the fastest-growing health challenges of the 21st century; with the number of adults living with diabetes has more than tripled over the past 20 years. In 2019, according to studies conducted by the International Diabetes Federation, the estimated number of people who have diabetes was 463 million and was projected to rise to 700 million in 2045. It is a major healthcare problem worldwide which, if not treated, will lead to other complicated medical conditions, including blindness. Early screening is highly effective in helping to prevent visual complications.
SELENA+ is an intelligent deep learning system that can detect eye diseases, namely diabetic retinopathy (DR), glaucoma suspects (GS), and age-related macular degeneration (AMD). Our solution is the perfect fit to resolve this onerous burden to screen an ever-expanding diabetic population amid increasing affluence worldwide.
Q. Besides the retinal diseases, are you looking at other new solutions or services?
Yes, we are in continuous engagement with inventors from SERI and NUS on AI applications to uncover systemic diseases. We hope to be able to announce some progress soon. As far back as the early 2000s, Prof Wong Tien Yin, one inventor of SELENA+ and current Medical Director of SNEC, had suspected a co-relation between retinal microvascular abnormalities and cardiovascular diseases and there is active research in this specific area.
At Singapore’s Ministry of Health (MOH) Committee of Supply Debate held on 5 Mar 2020, Minister for Health Mr. Gan Kim Yong shared, “Building on SELENA+’s capabilities, a predictive risk assessment model for cardiovascular disease will be developed to help doctors accurately identify high-risk patients and conduct more timely interventions to save lives and achieve better outcomes.” This is an indicator of the direction EyRIS is heading.
Meet the leader behind the success of EyRIS
Lai Teik Kin is a Co-founder of the Nova Group of Companies, focusing on development of software solutions for E-Government and Healthcare with implementations in more than 15 countries. The parent company Nova MSC was public-listed in MESDAQ Malaysia in August 2003. In 2018, novaHEALTH incorporated EyRIS as a joint spin-off with inventors from Singapore Eye Research Institute (SERI) and NUS School of Computing (NUS-SoC) to further develop and commercialize SELENA+, an intelligent deep learning system that can detect eye diseases. Teik Kin became the founding CEO of this start-up.