Collect, process, communicate – retina measurements with Mimo

Continuous monitoring is an essential process with every disease. In the case of eye disorders, frequent retina measurements can facilitate early detection of deterioration to quickly initiate intervention. This calls for comprehensive care settings, easy ways to take measurements and prompt results. However, in reality, this is rarely the case.

In this MEDICA.de interview, Dr. Peter Maloca talks about the shortcomings in the monitoring of eye conditions, explains how a newly developed device can solve this problem and describes the role artificial intelligence plays in all this.

Dr. Maloca, you have developed Mimo, a self-measuring OCT retina scanner. What prompted this idea?

Dr. Peter Maloca: We started researching in this area about fourteen years ago. At the time, I noticed that – unlike my colleagues –, I had always achieved better results in treating patients suffering from wet macular degeneration at my office. That's because I already had an optical coherence tomography (OCT) retinal laser scanner back then. It enabled me to immediately make the right diagnosis, quickly start the patient's treatment and closely monitor the disease course and effects of the therapy. Needless to say, the fact that OCT imaging systems were not comprehensively used for diagnostics and patients were frequently denied prompt appointments was a disappointing state of affairs we wanted to change. That's why we developed an OCT scanner that's fast and easy to use and facilitates measurements anytime, anywhere.

OCT has been used in ophthalmology for several years. What make Mimo unique and innovative?

Maloca: Mimo – that's the acronym for "mini and mobile" OCT – is not just a novel OCT scanner, but a self-measuring data system. This is what OCT tests normally look like: The patient must first take the time and visit the facility that has this type of devices. An assistant is needed to prepare and position the patient, take the measurements and save the data. Afterwards, the ophthalmologist typically reviews the images in the absence of the patient. He/she then has to schedule the patient for a return visit to discuss the findings. In collaboration with former Porsche designers, we have now developed a novel concept where the imaging device is more than just a machine. In this case, the patient embraces the miniaturized device like he/she would hug a person. To trigger the retina measurement, the patient puts his/her head on a fitted support brace. This sends an electrical signal to the device to start measuring. The device also detects if the patient slides back and is able to subsequently stop the measurement. The measuring process takes less than two seconds for each eye.

This process permits frequent measurements but the device generates a large body of data a physician is no longer able to analyze. That's why we have also developed a new machine learning software that analyzes the data. The results are communicated to the patient as recommendations and sent to a mobile device for example. Authorized persons such as family members or the ophthalmologist can be included in this process if the patient agrees to it. The human being always takes center stage over artificial intelligence in this setting. The ophthalmologist and his patient will ultimately choose a treatment together. That's why Mimo is part of a brand-new concept: First, we have to collect the data, then we have to process the data and third, the data must be quickly and effectively communicated. The data should flow and not be buried in a data graveyard.

For which disorders is Mimo best suited?

Maloca: Mimo is primarily aimed at patients suffering from age-related macular degeneration (AMD). The concern with these patients is that they stop coming to checkups as the years go by. Since our device enables us to take measurements without time or location constraints, delivers fast results and is easy to use, our hope is to keep these patients longer in the loop and encourage personalized ophthalmology. Having said that, the device is generally applicable for all conditions that involve an accumulation of fluid in or under the retina. Retinal monitoring is also an option with our system in the case of diabetes, vascular occlusion and inflammation for example.

What are some important points to consider with this application?

Maloca: One point is that this first generation device must still be used in a somewhat controlled environment. An expert should be present to ensure that the device is clean and check for dust or dirt particles or other external influences that might affect the measurement. That's why we suggest an application of the device under supervision – at an ophthalmologist's office, at a hospital, retirement home or optometrist.

Another point to consider is that patients must be prepared to take measurements on a regular basis. Some AMD patients are unwilling to undergo any type of intervention. This is very dangerous, especially since the likelihood of conversion of the other, previously non-diseased eye increases with age. After the age of seventy, the risk increases by nearly twenty-five percent within five years. In theory, our device could not only monitor the affected eye but also show an early conversion of the other eye long before the patient even experiences any symptoms. Studies have also shown that regular monitoring makes earlier interventions possible and prevents vision loss. What's more, the large amount of data supports the decision-making process, which in turn also justifies the cost of treatment.

In your opinion, what role will artificial intelligence play in the future of medicine?

Maloca: The future has already begun in this case. We are already in the midst of it! When you use your credit card, stream a video or shop online – artificial intelligence is already a part of it. However, we prefer to refer to it as machine learning because these methods are simply advanced calculation methods. All of these algorithms have one thing in common: they do an excellent job within a very narrow and restricted framework. They are so exact, fast and efficient to where human beings are both very impressed and also deeply afraid at the same time. Our research aims at reducing fears and proving that humans and machines can also work together – and in such a way to where human beings always must make the ultimate decision. By visually and mathematically illustrating the results, the software allows us to understand the test readings and continue to learn new things. Thanks to its reliability and efficiency, physicians save time and are free to focus on other important tasks – namely, serving their patients. Machine learning helps us expand our expertise and improve the physician-patient relationship.