Startup Spotlight: Saccade Analytics™ – Pioneering Eye-Head Movement Tools to Aid in Diagnosis of Over 200 Diseases

Editor’s Note: McGill Dobson Centre Ambassador Nely Gaulea sat down with Saccade Analytics™, this year’s McGill Dobson Cup 1st place winners in the Health Sciences Track, just prior to their win announcement. With all due congratulations, it is clear that this team shares vested passion for a common goal, a big vision, and an even bigger purpose. Learn more about their scientific discovery and the onset of their entrepreneurial journey building on decades of groundbreaking biomedical research at McGill University.

READ ALSO: Winners of the McGill Dobson Cup 2017

Dr. Henrietta (Mimi) Galiana

Dr. Henrietta (Mimi) Galiana, professor in the Biomedical Engineering Department, is McGill University’s very own pioneer in model-based analysis for eye and head coordination. A distinguished academic and previously, the only female student in the McGill Faculty of Engineering entering class of 1961, Dr. Galiana’s decades of research have set the groundwork for Saccade Analytics. The startup born in 2016 capitalizes on scientific insight and the infinite possibilities of emerging eye tracking VR technologies to provide revolutionary quantitative metrics to aid in diagnosis of over 200 diseases with traces in eye and head movement such as concussion and dizziness.

In addition to Dr. Galiana who serves as president and scientific director, the dream team behind Saccade Analytics (pictured in the cover photo above) consists of CEO Dr. Iman Haji, PhD’15 and post-doctoral fellow in Biomedical Engineering; COO Isabel Galiana, PhD candidate in Economics; product developer Heather Armstrong, M.Eng. in Biomedical Engineering; and Dr. Francisco Galiana, Emeritus professor in Power Systems and Electrical Engineering at McGill University, who recently joined the team as R&D Lead and chief inventor.


About Saccade Analytics™

Saccade Analytics™ provides neural disease diagnostic solutions in the form of automated software, to analyze eye and head movements and generate health metrics based on them. These metrics help doctors to diagnose diseases and to evaluate the progression of recovery. The numerous neural diseases that influence eye-head movements include and are not limited to concussion and dizziness.


How did it all start? What made you take the leap into entrepreneurship?

IH: I did my PhD in Biomedical Engineering at McGill under Prof. Galiana’s supervision. The topic of my research was understanding how the brain works as it pertains to eye-head movements. Prof. Galiana (Mimi) is “the” pioneer in this area. She shared her vision with me and I was quite impressed. During my PhD, I expanded on those ideas and we came up with the most advanced model-based analysis for eye and head coordination in the world. We published it in the Journal of Neurophysiology a while ago. It was pure research at the time and that was the work of my PhD. Given this model, we realized that we could do many things with it and the insight that accompanies it. In fact, Prof. Galiana had this unique understanding for decades.  We were at a stage where we were able to demonstrate it to everyone. We merged knowledge from neurophysiology, behavior and implications for diseases that this gave us a unique viewpoint. We soon realized that we could actually help people and we had something to share with others.

We came up with the most advanced model-based analysis for eye and head coordination in the world.

After my PhD, we decided to use our acquired insight to design diagnostic tools and protocols. We needed tools to extract diagnostic information from eye and head movements. Why is this important? Doctors measure your heartbeat rate, your temperature, and other metrics from your body… but currently, eye and head movements are not considered as widely as they should be as sources for diagnostic data. Particularly, the eyes are very important because they move very fast, quickly jumping around… They say that “the eyes are the windows to the soul” but they don’t use them in diagnosis! So, there is potential here. This was the basis for the formation of Saccade Analytics.

MG: The logic is that as you learn a particular piece of the brain, how it works, and how it’s connected – you understand how the responses in different places change under different conditions. That allows you to make up a model. A model is nothing more than a schematic and some math that embodies everything you know about the system.  I realized for many years that analysis had to change, but nobody stepped up because they couldn’t do the necessary mathematical work. Traditionally, recordings in clinics based on eye and head movements produce noisy results, because the analyses are wrong. They are measuring the wrong thing and on top of it, the wrong way given the neural connectivity. They are making assumptions that don’t satisfy how the system behaves. So, when we design new algorithms and analysis methods, we can extract much more robust information, more accurate information. Furthermore, it can relate to particular sites in the brain and so it also helps the doctor localize lesions whether at the eye level, or the sensors in the ears, or in between. To start, we’re working mostly in vestibulo-ocular reflexes (VOR) – the basic function related to dizziness which also involves eye and head coordination. That’s where we have the most knowledge and so, where we can have the quickest impact. Our models act as testbeds (surrogates for the brain) on which to validate new analysis tools.

Saccade Analytics™ models act as testbeds on which to validate new analysis tools.

Part of the decision [to start this venture] was because, despite my recommendations to colleagues who really liked what I was doing, they wouldn’t adopt the new way of measuring things because they said it was too complicated. “If you want us to use it, you’ll have to build it and develop it for us.” So, I knew that it would never happen unless I – at least – started it.

IG: From my perspective, what made it a viable business for me, is that if you look in the literature, there is a strong correlation between eye and head coordination and diseases. There is also an unwritten consensus that innovation in data collection and analysis is overdue. We all agree that some information is lost with current methods, but now we could recover it.

Tell me, how are you addressing this problem considering the current state of the art in healthcare?

IG: The fundamentals behind most protocols are mainly subjective visual tracking [e.g., visually following the doctor’s finger] and answering some general awareness questions. We’ve only tested a few concussed patients by actually recording their eye and head movements on a PC. Hence, we can compare numerical metrics during subsequent tests. Previously, patients had little information on their concussion recovery over time. With us, they keep asking if they can come back for re-tests because they want to know if they are getting any better. They go to their doctor but their doctor has little to offer quantitatively, short of fMRI scans.

HA [a four-time concussed patient herself]: Basically, doctors ask you questions; there is no physical measurement. The only measure right now is “try this [exercise] and if you feel really bad the next day, that was probably too much.” Literally, it’s trial and error towards recovery.  

MG: It’s a clinical interaction. Nothing is measured, really. The other thing is that – and concussions are a good example – they have ad hoc subjective criteria with clinical practice, but then, they just send you home.  They instruct the patient to come back in two weeks or a month, and in-between there’s nothing, no contact. There’s nothing to guide the patient if they are getting any better or if they are doing something wrong. With the new mobile recording devices, through eye-head movements, we can actually give them something with a click of a button on their phone eventually saying: well, “yes, your eyes and head are better coordinated now and you’re getting there, you’re 50% there,” or something general, before they see the doctor, or “stop,” or “don’t do this, it’s triggering your problem.”

It’s a similar approach to what a cardiologist does. He sends you to the pharmacy to rent a heartbeat and blood pressure monitor. The recordings are stored in the device and they look at that when they see you the next time. The same approach should be possible for concussion and dizziness. The reason it wasn’t done in the past is because the analysis couldn’t be done except with a human operator applying the algorithms and that required a central site. Now, with the new portable computers and software, you can design apps that can do analysis on the fly, no problem. Another major reason I was interested in this concept originally, was that if someone needs testing and certification in a skilled occupation, their competence [lack of dizziness or concussion sequels] should be evaluated in the context of their normal activities! For example, if you’re testing a track athlete, you want to test them while they’re running not while they are sitting on a chair in the clinic.

Current eye-head recordings in clinics require large expensive equipment and tests can take all morning, and over a day for analysis. With Saccade Analytics™, it’s possible to have results in one minute.

Doctors have limited time. The current eye-head recordings in clinics require large expensive equipment and tests can take all morning. The technician will take over a day for analysis, so you can’t do too many tests in a week in one hospital. With our tool [that comes in the form of automated diagnostic aid software], it’s possible to have a result in a minute and the measurement can be done the moment the patient comes in clinic, not 2 weeks or a month later. That is a long time when you can’t stand up and you’re feeling dizzy and uncomfortable. It’s very difficult for patients. We attended talks by athletes who had concussions and the biggest problem was their fear that they were going crazy. They didn’t feel right and they had nothing to tell them if they were getting any better. One of them broke into tears describing how he felt…

IG: One alternative right now is the fMRI which is expensive. But then again, this is what happened to one of our recent test subjects: she was not feeling well, she went to the doctor, she had an fMRI, and the results were inconclusive. There was nothing there. Later, she got positively diagnosed with a concussion by a physician. Yet, [with our tools] even a month after her concussion we could still clearly see dysfunction in her eye movements that correlates with concussion.

IH: The fMRI doesn’t give you much. The thing is that when you do imaging scans, they don’t accurately evaluate your behavior, your movements, and so on… It’s basically looking at neural connections and neural firing rates, albeit with a delay. The current state of knowledge cannot interpret how that translates into movement deficits with a great accuracy. There is ongoing research on this, of course. In brief, fMRI only shows firing activity of neurons and not movement deficits – the latter can only be detected by actually measuring movements.

What is your bigger vision?

IH: Over 200 diseases have traces in eye and head movements. These diseases are not limited to dizziness and concussions. Many examples are reported is psychiatric problems, even neurodegenerative diseases such as Alzheimer’s. Assisting diagnosis in more diseases is our bigger vision for the future.

IG: The bigger vision is Prof. Galiana’s vision: to have this wearable device and test people in their daily lives and activities so that we can help support diagnosis, monitoring, early diagnosis of neurodegenerative diseases, determine if treatments are working or not, and all of this from a person’s home.

MG: To help people and to give them measures that are relevant for what they do so that they can live their normal life.

IH: To determine new eye-head movement metrics and make them the standard for helping doctors diagnose numerous diseases, and adjust treatment during recovery if necessary.

Over 200 diseases have traces in eye and head movements. […] Assisting diagnosis in more diseases [including psychiatric problems and neurodegenerative diseases such as Alzheimer’s] is our bigger vision for the future.

You were part of the McGill Lean Startup 2016 cohort. Why did you decide to participate in this program and what were your key takeaways?

IH: We had a provisional patent on an algorithm developed during my PhD and Derrick Wong, our technology transfer manager [McGill Faculty of Science and Engineering], suggested the McGill Lean Startup program to us last September. He said it was a program for those interested in growing their business. We acknowledge his support in this endeavor!

IG: The McGill Lean Startup really helped us compartmentalize all the different aspects, lose the fear [about embarking on an entrepreneurial journey] and identify our market segments better. At the beginning, we thought that we could use our tool for everything and that we could solve all the problems of the world! The McGill Lean Startup program helped us focus our efforts on specific market segments.

MG: It helped us to focus. Now, we know that the different targets will need tuning for what they want. It means treating different market segments individually. We now have a solution tailored for dizziness clinics, and the whole exercise showed us that if we compartmentalize our target markets, the whole process becomes manageable, so we gave it a try.

READ ALSO: McGill Lean Startup 2016 Final Presentations Recap

What about your experience in the McGill Dobson Cup 2017?

HA: What you do in the McGill Lean Startup leads into what you need for the McGill Dobson Cup. We were motivated by the McGill Lean Startup and the McGill Dobson Cup seemed like the logical next step for us. It was a chance to take our venture a step further and see how we stacked up against other teams.

IG: Writing the 5-page business plan was helpful. Also, in the McGill Dobson Cup semi-finals and the finals, the judges themselves provided new viewpoints to polish our pitch. Since they were selected with expertise to match their assigned track [in this case, Health Sciences], their comments were very pertinent. In fact, one of the judges in the semi-finals, Dr. Brent Norton, has helped us out. Another judge from the finals has shown an interest in investing. The fact that it was targeted was very useful for us.

From the business side, because it’s a medical device it’s a little more complicated. We have to deal with regulatory issues which can be difficult. Once we start looking more intensely for partners, hires or for financing it will get even more complicated, but we are well surrounded with our new mentors and we have a good structure around us.

Saccade Analytics™ in the McGill Dobson Cup 2017 finals – March 22, 2017

Let’s talk a little bit about your pitch. I have seen you pitch three times already: at the end of the McGill Lean Startup program in December, and more recently, in the McGill Dobson Cup 2017 semi-finals and finals. Your pitch was better and better every time. How did you go about improving it?

IH: The McGill Lean Startup program director, Renjie Butalid was really, really helpful in terms of pitching. We had at least 25 iterations of our pitch. We also got different viewpoints from other people that he recommended that also helped us refine our pitch. And we’re still doing that… Every time we pitch – and you can testify to that – it gets better. I would like to acknowledge someone in particular who coached us for the McGill Dobson Cup 2017 final pitch: Alexander Haque, CEO of Retinad VR and former director of the McGill Lean Startup program. Despite the fact that he was very busy, he invited us to his office to pitch, gave us really good feedback and comments from his own experience, and told us to come back in 3 days and pitch again. Which we did. You can imagine it was difficult to incorporate all the changes in such a short time, but we did it and it made a huge difference.

HA: One of our communication challenges is in translating from the academic world, where there is a lot of emphasis and debate around technical details, to the business world. People are genuinely interested when you’re starting a business. It’s exciting and they want to know more so we get to constantly explain what we do. It’s been really helpful to think about these everyday interactions and how we explain things casually to friends or family when we’re not focused on providing a formal or structured explanation.  This helps us see what people understand or what they get excited about and then we try to work these aspects into our pitch.

What have been some of the lessons learned throughout your entrepreneurial journey thus far?

IH: We learned how to formulate a product to market fit. We learned about product iterations and connecting with the target market very early so that you are not designing an optimal solution that nobody wants. We have applied in practice everything we’ve learned. For instance, we did a few demos for clinicians at the Glen hospital. It sparked the clinicians’ interest in our product and in collaborating with us. The idea of going out there very quickly and not staying confined within your office walls was the main point for me. The lesson was to avoid trying to generalize a solution and instead find out what a target audience really wants or needs.

MG: Something that I learned that I hadn’t learned in all the years of research, is that scientific lingo can be a pain in the neck for the general population, and even for many researchers. When they didn’t understand me in a conference or didn’t believe me, it was because of the way I said it. When I changed it, the door opened, just by changing the dictionary that I used. That, to me was a big element that was very positive.

HA: Since we started our business, we constantly must explain to others what we do. It requires checking what they understand or don’t understand and clarifying as we go. Academia, to a large extent, is very detailed and if you’re doing a PhD or a master’s, it’s focused on one topic. Whereas, in a business, we’re doing something that is generally applicable so we must find a way to explain how it’s practically useful to the layman. People are genuinely interested when you’re starting a business. It’s exciting and they want to know more.

Saccade Analytics™ in the McGill Dobson Cup 2017 semi-finals – February 14, 2017

What has been the most valuable piece of advice from your mentors?

IG: The first point is to get out there and ask questions. Ask if what you’re doing is useful, ask them to clarify what they want, how it could be improved… Don’t sit and try to perfect something if you have no idea what people want. Find out if your product is wanted and how you can tweak it. The other main point is to avoid repetitions about your current product; instead, talk about the long-term vision which is much more exciting. So, I think you have to talk about the vision and get out there and find out what people want.

IH: Talk to as many people as you can. It really works! Once you explain yourself to them and ask them, “does this make sense to you or not?”, the feedback you get is crucial. We get new feedback everyday just by talking to other people: those who we pitch to, and those who we want to target as customers.

What resources have proven to be the most useful to you as new entrepreneurs?

IH: The entire course in the McGill Lean Startup program was helpful; the Lean Business Canvas, in parallel with the online course Lean Startup Methodology by Steve Blank recommended by Renjie were very useful. At the same time, we audited the course FACC 500 Technology Business Plan Design by Prof. Michael Avedesian. Fantastic course! We want to thank Prof. Avedesian for allowing us to audit this course, for his suggestions on our journey and also for inviting us to take part in the course’s workshop FACC 501 Technology Business Plan Project. He has been very helpful and kind.

MG: In other words, there was a counter balance of resources that were flagged, that we didn’t even know about and that were made available for us right here at McGill.

What’s next for Saccade Analytics™?

IG: Right now, we have two clinical trials starting. Soldier medical testing with the Canadian and U.S. military. These are our focus over the summer.

We are also applying for the McGill X-1 Accelerator. We would like to all be working from the same space. At the moment, we are all scattered and that’s really difficult. I think that the accelerator will be more specific than the McGill Lean Startup and the Demo Days would be a great exposure. Hopefully, that will lead to Demos in Boston and in San Francisco.

IH: We do have a beta product and we are ready to accelerate. Once we have enough data, we’ll refine our algorithms and improve our metrics. From there, we’d like to expand… To go mobile. We’d like to be present in general practitioners’ offices and make this part of annual check-ups to diagnose problems such as neurodegenerative diseases in their very early stages.

HA: In the McGill Lean Startup program last fall, we were just getting started. Now, we have something more substantial and it’s time to take that out into the world.

Anything else that you’d like to add?

IG: If anybody is out there with a good idea, get in touch with the McGill Dobson Centre for Entrepreneurship! It’s super helpful.

To learn more about Saccade Analytics, check out their website at saccadeanalytics.com.


Thank you for this very insightful interview and congratulations on your McGill Dobson Cup 2017 win!

mm

Nely Gaulea

Nely is a graduate student in Strategy & Entrepreneurship and an ambassador for the McGill Dobson Centre. Her main interest is in the life sciences sector and collaborative innovation at the intersection of business, healthcare and technology.