Skip to main content

Retinal imaging prevention and early detection of chronic diseases

Periodic Reporting for period 2 - RETINAL (Retinal imaging prevention and early detection of chronic diseases )

Reporting period: 2019-06-01 to 2020-05-31

RETINAL project is part of a broader initiative in the area of health care and prevention of Non-Communicable Chronic Diseases (NCDs), based on the scientific evidence that some systemic diseases cause changes in the micro-vasculature of the retina. The retina is the only part of the human body where blood vessels can be observed directly in a non-invasive manner by examining the ocular fundus.

The idea under this initiative is to facilitate the early diagnosis of diseases by examining the anatomical structure of the retina and to enable timely management and effective treatments, having a great impact in the individuals’ health and wellbeing and saving a great amount of money to health systems and the overall economy. For this purpose, ULMA Embedded Solutions (UES) has developed non-invasive and cost-effective ICT tools (U-RETINAL) for automatic quantification of retinal parameters and automatic detection of lesions of the retina, which can be used in health interventions delivered through primary health-care centres to strengthen early detection and timely treatment of various diseases. U-RETINAL is a comprehensive solution for ophthalmic imaging analysis composed of:

1) A simple, reliable and cost-effective digital fundus camera (U-RETINAL DFC)
2) An extensible set of software modules for specific clinical applications:
• U-RETINAL AVR is already available, but only for research purposes, since more scientific evidence is needed before its adoption in clinical practice (to screen population for early detection of NCDs).
• U-RETINAL DR is already available and ready to be introduced into the market.
• Other modules will be developed in the short-term.

Our solution is based on:
1) The U-RETINAL Digital Fundus Camera, which is economic, easy to use, portable, and non-mydriatic. These features make possible:
• To perform the screening for Diabetic Retinopathy (DR) in the regular visits of patients to the physician for diabetic care, because U-RETINAL DFC cost is affordable for primary health centers, and it is reliable and simple to use by general practitioners or nurses.
• To overcome some patients concerns about the potentially adverse effects of pharmacological drops and ensure a better compliance of patients, since U-RETINAL DFC is a non-mydriatic fundus camera, which means that no pharmacological drops are needed for capturing the retinal images. Non-mydriatic retinal digital photography has demonstrated a good performance for the purpose of DR screening.

2) The software module U-RETINAL DR for automatic detection of the presence of DR, which allows:
• To make DR screening cost-effective and more agile.
• To provide immediate feedback to patients about the presence or absence of DR, which is another important factor to achieve high rates of DR screening follow-up.

Our ambition as a company is to change the current paradigms in the clinical/medical practice through the development and introduction into the market of ICT tools for the early diagnosis of several diseases which cause changes in the microvasculature of the retina, with the aim of improving efficacy of treatment of chronic diseases that affect a wide population worldwide, and therefore improving health and quality of life of patients. “We pursue the creation of new paradigms in medical practice for breaking current barriers”. Our goals are:
1) To modify current manual systems in medical practice by introducing ICT tools more accurate and efficient.
2) To gain the recognition of the ULMA Trademark in the medical sector, as a high-qualified technological provider which exceeds customers expectative.
3) To be a market reference in Europe by achieving a high impact in the efficacy and efficiency of DRS Programmes. We pursue that 70% of diabetic population are screened for retinopathy with our DR screening tools in the next years in Europe.
4) Our goal is to enter other markets taking advantage of our position in Europe and to achieve similar impact in the health systems in those markets.
During the course of the first and second year of the RETINAL Project, UES has been devoted to developing the following works and actions related to the development of the RETINAL Solution:
Current limitations of DR screening processes and derived business opportunity:
On one hand, grading requires personnel with suitable skills and experience. Graders need to gain experience in disease identification by assessing sufficient examples of the various clinical indicators of DR. The quality standards stipulate that graders should be grading at least 1000 patients per annum (and up to 5000) to ensure that they see the whole spectrum of disease sufficiently frequently to maintain expertise. On the other hand, grading takes time and it is generally considered that grading more than half of the working day might be too tiring for an individual.
To avoid this dependence on the skills and experience of the grader and to minimize the time of the screening process, it is hoped that stage L1 GRADER (disease / no disease grading) will be performed by automated grading systems. Additionally, given the increasing prevalence of diabetes worldwide, national screening programmes have difficulties to meet the ever-increasing demand of patients by using manual grading alone. Automated grading, therefore, offers a means to improve the efficiency of national DR screening programmes. Some software packages seem to be getting reasonable results for the detection of microaneurysms and some other features of background DR (haemorrhages, exudates). Software must be capable of identifying securely the presence of any type of retinopathy. Although it is recommended the use of automated software for distinguishing retinopathy from no retinopathy, currently there are a few commercial solutions available and they are not widely introduced in the market: they are oriented to screen customers in optometrist chains as an added value service, or to reduce the workload of graders in some Grading Centers. Its usefulness and acceptance depends on the following key factors:
• Adequate levels of sensitivity and specificity for detecting retinopathy
• An appropriate technical failure rate: ungradable photographs because of insufficient quality
Our business opportunity derives from this demand. U-RETINAL DR software achieves good results in the detection of the presence of any kind of DR, and therefore it is useful for the purpose of automating the first stage of the DR screening process:
• Sensitivity: 91,6%-94,8% (sensitivity ensures that disease is not being missed)
• Specificity: 80,2-84,7% (specificity minimises inappropriate referrals, and associated patients’ anxiety)
• Technical failure rate: 5,2% (this rate is between the interval 5%-10%, which is considered appropriate)
Our solution prioritizes sensitivity for detecting retinopathy (not missing cases of DR), although this can be associated with lower specificity (more referrals).