SINGAPORE (Jan 14): Virtual reality and its applications tend to be associated with electronic games and entertainment. But the immersive technology has increasingly been deployed in various other fields, particularly medicine. A Singapore-based start-up, for instance, is using VR headsets to make the diagnosis of glaucoma easier. BetaSight’s solution tracks a patient’s eye movements in response to pop-up virtual objects generated by the VR headset.
Glaucoma is an eye condition in which the optic nerve is damaged as a result of abnormally high pressure in the eye. If the damage continues, it can cause permanent vision loss. According to BetaSight founders Martin Sawtell and Corey Manders, the current process of diagnosis can be taxing and lead to skewed results. One of the start-up’s investors even had to take the test twice, and was charged double the price.
The current procedure, using a visual field analyser, requires patients to remain completely still while staring at a spot of light at the centre of their field of vision. Patients will have to indicate, usually by pressing a button, whenever they see a pinpoint of light flash in their peripheral field of vision. In some cases, those undergoing the test may be anxious and inadvertently skew the results by pressing the button regardless of whether they see a flash of light.
“It’s going to mess up everything,” says BetaSight’s Sawtell in an interview with The Edge Singapore.
With BetaSight’s VR headsets, however, patients do not have to sit still or press any buttons. Instead, they only have to put on the VR headset and track the virtual objects with their eyes, which is more intuitive than pressing a button. A person who has glaucoma will have a diminished field of vision and not be able to spot all of the virtual objects. The VR headset is portable and considerably less bulky than the visual field analyser equipment in use today.
Manders says BetaSight has already received approval for clinical trials, and the technology has been tested on 350 people via focus groups. These have so far garnered positive responses. But, he adds that “we are looking for more [people to undergo trials]”.
Lightbulb idea
The idea of using VR technology to diagnose glaucoma first came to Sawtell when he visited a doctor after suffering from a bout of vertigo. The doctor examined his eye and, based on the lag in focus from his eyes, was able to diagnose that he had a fracture in his skull.
The episode led Sawtell to think about the possible applications deriving from eye check-ups to diagnose medical issues. He had also read about eye-tracking headsets. At the time, Sawtell had joined start-up accelerator Entrepreneur First. There, he met Manders and they jointly pitched the idea to start BetaSight.
“[While] Entrepreneur First takes you in, you don’t need to have an idea [at the beginning]. But you need to [generate] ideas with other people within three months or you are out. Within a further three months, you need to validate your idea, if not you’re out. That’s what we have managed to do,” says Sawtell.
Sawtell had previously worked with Lucasfilm Singapore, the local subsidiary of US movie production company Lucasfilm that was acquired by The Walt Disney Co in 2012. His 10-year career in the film industry has turned out to be especially helpful in getting BetaSight off the ground. At Lucasfilm, he was involved in producing digital effects for movies, especially in the area of dynamic simulations, such as the movement of fluids and fabrics, and even explosions. “Representations of reality is my speciality,” Sawtell says. “I can figure out how to do it via algorithms. So, whatever you want to build, I can find a system to do it and then render it to make it look real.”
Manders’ experience has also been helpful in founding BetaSight. He was a research scientist at the Agency for Science, Technology and Research, or A*STAR. He had worked there for 10 years on various image-processing projects, including VR and augmented reality (AR). After A*STAR, Manders had a stint in a start-up that became successful, which gave him the desire to start his own business. He has also collaborated with the Singapore Eye Research Institute to develop a system to diagnose diabetic retinopathy.
Both Sawtell and Manders do not possess any medical training or expertise, so they rely on several ophthalmologists for advice. “These advisers participate in product development too. We have ideas and we need to validate them,” says Sawtell.
Leveraging data
Now, with BetaSight, Sawtell wants to “build something new” and different from what he has done so far. He believes VR and AR technologies hold enormous potential. “I have a massive passion for VR and AR. I really think they are going to be the future of computing, that is, spatial computing,” he says. “That is something that is here to stay.”
Indeed, Sawtell expects BetaSight’s VR headsets to eventually do more than diagnose glaucoma. He says eye movement data collected by the headsets could be used to diagnose other eye and neurological diseases. The key is in understanding how and why a person’s eyes move in a certain manner. “The vector of the eyes can show us more than just glaucoma,” he says. “We get insights into other diseases: isolated nerve dysfunction, diabetic retinopathy and potentially neurological diseases.”
For instance, eye movement data that positively correlates with patients who have attention deficit hyperactivity disorder can help improve future diagnosis of such conditions. This could also be extended to the diagnosis of brain diseases such as dementia, Parkinson’s and Alzheimer’s diseases and so on. “[We want to] look at cognitive processes from the way you move your eyes,” says Sawtell. “How does that reflect what you are thinking about? And what does that say about you?”
The critical step is to collect as much eye movement data as possible. Manders says BetaSight needs to conduct many trials to construct medical templates, which are necessary to identify correlations. “From eye-tracking data, we train the neural networks to discriminate between different eye diseases. So, it is like a deep-learning approach to classification as we conduct more and more tests,” he says.
“The headset enables a visual test field. We can detect if you are not seeing things in a certain place. It could be glaucoma, stroke or other reasons. We do know of people [who have] blind spots and are either aware or unaware of it or unsure why they have it. It is up to the doctor to examine further. But we can measure that. Our point is to show that you can do more with the data,” says Manders.
He adds that the company is focusing on gaining traction and recognition for the headset as a diagnostic tool first, and has kept the testing procedures simple and similar to existing protocols. But in the future, BetaSight could also inject fun into the process, by incorporating games into the headset, for instance. “Our goal is for you to be able to play a game and we’ll tell you if there is something wrong with you,” says Sawtell, noting how Apple’s smartwatch now comes equipped with an electrocardiography function that measures the wearer’s heartbeat rate in the background. This is because the trend in wearable devices now is to incorporate healthcare functions, he notes. “I’m very excited to look into this area.”
