TEXT: 415-425-4961

S1:E4 | #4 – If Robotic Surgery Only Had A Brain w/ Kris Siemionow, MD, PhD

UMN Kris Siemionow | Robotic Surgery

Dr. Kris Siemionow is a board certified orthopaedic surgeon. He completed orthopedic surgery residency at the Cleveland Clinic Foundation and spine surgery fellowship at Rush University Medical Center. Dr. Siemionow’s PhD focused on the effects of inflammation on nerve cell function.

Dr. Siemionow is the Co-founder & CEO of Holo Surgical, a company developing a proprietary neural network and algorithm to deliver the world’s first-in-human digital surgery based on augmented reality and artificial intelligence.

Listen to the podcast here


If Robotic Surgery Only Had A Brain w/ Kris Siemionow, MD, PhD

We have a very special guest and a good friend of mine, Dr. Kris Siemionow. He is a board-certified orthopedic surgeon and entrepreneur who has founded multiple healthcare companies that utilize advanced technology to improve clinical outcomes. Kris completed his residency at the Cleveland Clinic Foundation in Orthopedic Surgery and Spine Fellowship at Rush University Medical Center.

He has Ph. D. focused on the effects of inflammation on nerve cell function. He is the Co-Founder of Global Spine Outreach, a nonprofit whose mission is to save children with severe spinal deformities. He also Co-founded a number of other companies. We’re going to let him talk about it. Kris, you got exhausted doing all of the things you’ve done already. You are still a very young guy. First of all, welcome and thanks for being on the show.

Thanks for having me. As you probably are aware, physicians love to collect titles and there has to be a long list of accomplishments behind every physician’s name. That’s joking aside.

When did you move to the United States?

The first time I was in the US is in 1986. My mom was doing a hand surgery fellowship at the Christine Kleinert Institute in Louisville, Kentucky. I came on a cold December day from Eastern Europe to Kentucky. In 1986, it was a pretty big culture shock. We came back to Poland, which is where I’m from, where I was born and raised, and then came back to Kentucky again in 1989. It’s a little bit of back and forth, and then I spread out all over the United States from that point forward, essentially moving from state to state until I was in medical school.

In 1986, you were there in Kentucky for three years.

We spent about two years in Kentucky. My mom did a hand fellowship, then we went back to Poland When the world was opening up with the Berlin Wall and everything coming down. 1989 was when my parents decided to come to the US permanently.

When you moved back to Poland, were you excited and you see your friends, or are you hoping you would stick around in the US?

You’re a kid. They put you on a plane. You probably don’t even know what’s going on. You accept things for what they are. Children are more plastic. They adopt a lot easier. I had friends and family I was coming to. It was not a big deal, not that I recall.

Once 1989 hit, you were in the US.

The world was moving everywhere. You have followed the Eastern Europeans and Russians. Everybody was coming to the US and there was a big wave of migration around that time. We were part of that.

These things that happen in history that are unexpected or maybe hoped for, like the fall of the Berlin Wall become catalysts for change. In your case, early on in your life, Berlin Wall falls. How did that influence your parents’ decision to move to the US?

It’s no secret that living under any communist regime is challenging. A fair few at the top were sticking around, but for most of the society, the biggest benefit is in leaving. My parents, in a similar fashion, found much more opportunities. They’re both scientists and then they felt that they would come to the US and be able to pursue their academic careers, which did happen. They did very well as a result. Most people that came in that same era also probably did not have too many second thoughts professionally. At least, there’s a price to pay in the sense of leaving family and friends behind which is what you carry with you. Interestingly, what you say about COVID and there being impact on you counts during our lives.

The most challenging times teach people to get by, live with the problems, and adjust accordingly. Click To Tweet

How many of these very rare occurrences have happened? It’s astonishing to me. We’ve had 9/11, the 2008 market crash, the dot-com bubble, and the Berlin Wall. Every decade, we have some major events that never happened before. We learn to get by, live with it, and adjust. Every period after each one of those events that I mentioned has been more positive and better. I’m counting on that happening here and then moving forward.

You’re kind enough to share a book recommendation with me and I try to send one back. You read probably five for every book that I read. One of the books we talked about is The Black Swan Event. The point you made is interesting because, in my career, I’ve been working for many years. Dot-com was the front end of that. 9/11 came right after that and then the 2008 financial crisis. If in fact, we’re heading into a COVID-19 recession, which some people think we are, that is a lot of black swan events.

That doesn’t make them black swanish anymore.

In that case, as we think about our careers, plans for what we want to do, and the projects we tackle, it’s almost like not if, it’s when and how is your plan going to be nimble enough to first endure the unexpected event. While everybody else is worried and retrenched, maybe you can advance and gain some headway. The Vision Fund, which is a part of SoftBank, presented its quarterly earnings. For the audience that doesn’t know, they raised the largest venture capital fund in history. It is $100 billion. They’ve now deployed all of that capital. If you take the number of days since they had the funds, since they closed, they’ve averaged $100 million of investment per day.

What I’ve found fascinating is they have this great slide deck. There are all funny slides like unicorns falling into cliffs, but there’s a slide that compares the degree of innovation and technology that drove a large part of the US economy for many years after The Great Depression. After the COVID Recession, there were six blocks. This is like visualizing where they see the investment themes and the biggest opportunities.

Although the Unmet Need is focused on healthcare innovation, that’s not what the vision fund focuses on. They focus on Uber, WeWork, or any type of technology business, but out of six themes coming out of COVID-19 telemedicine, telehealth is one of them. That’s probably going to be one of the areas with the most growth. Regardless of the line of work we’re all in when everybody has to stay home for months to prevent waves of sick people coming into a hospital infrastructure that can’t withstand the waves, healthcare is front and center on everyone’s mind.

It coincides with going back to 2001 when the dot-com bubble happened and 9/11 shortly after. Most of the underlying technologies that are enabling a lot of telemedicine telehealth, most of the important things that drive our economy now were just getting started. People getting broadband. It’s interesting that we would start there. With that background, you get to the US. How many years until you settled in and said, “This is home. I like it here.”

I have the advantage of coming as a kid. It’s so much easier. When you come in your late teens, it becomes harder. When he comes in my father’s age, who was in his early 40s, it becomes difficult. For me, it was within a week. My English wasn’t great, but I was definitely feeling that I can do everything I want to do. That was good. There were kids like me playing and that’s most important when you’re 8 or 9 years old.

Both of your parents have academics. Did you have a natural interest in school, science, and learning, or was that something you developed over time?

Absolutely not. I don’t think I was a great student. I was not interested in what was happening academically in school. I had my own interests, which were nerdy, but not in line with the school curriculum. I was into like reptiles and lizards. I’m machinery associated with tanks, and planes, and would try to spend a lot of time on that, researching, and reading. I’m doing things that you may consider productive but not by school standards. It was a challenge for my parents to get me interested in school.

Did they see at least the curiosity that you’re exploring all these topics or did they want you to be more focused on getting As?

Like any other parent, they judge you like the school judges you for most of it, but they themselves are probably not super aware of the school system here. In Poland and Europe, the schools are mostly focused on memorizing telephone books and impractical things, and then you have to recite them which is silly in the days of Google Assistant and Alexa, where you can get all that information. You don’t even need to know how to read and write. You just talk. That’s how school is there. If you are not able to produce that content, then you’re probably not impressing many people.

I had my own interest. I stuck with them and things ended up okay. I’m seeing similar interests in our kids. You got to support them because forcing them to do things is not a good way to go here. Knowing that it ended up okay for me or other people that were doing similar things is encouraging. Not that I’m doing some Ph. D. in education, but I think that’s common sense. If you’re interested in something and those pursuits can translate and morph into something more productive in the future as you grow.

UMN Kris Siemionow | Robotic Surgery
Robotic Surgery: Forcing your kids to do things is not a good way to go. Just support and always encouraging them to do the things they like.


Curiosity and being encouraged to explore things you’re interested in for the love of the information. That’s something I value as part of my education. I try to recognize that when my kids practice it versus I have to memorize this phone book so that I can get a 99%, 100%, and A-plus. Was grade something that validated you when you were in school?

Only if they would have been good. Most of the time, they were not. In grade school and even early high school, I did my own thing. It’s not that I wasn’t paying attention. I was unfocused. It wasn’t something that was interesting to me. It’s got like a late bloomer scenario. I became interested in things that would make teachers and parents happy probably in my very late teens or early twenties. I became interested in organic chemistry when I was 20 or 21, which was a good time for what I was trying to do. Before that, it was not something that would find much time for me.

That’s the problem with the educational system. Everybody develops at a different pace and sometimes there’s not enough time, room, or runaway for people that are maybe late bloomers or find those interests later in life. The nice thing about America is that there are many avenues to get into things. There are plenty of people that switch careers. There are plenty of those histories. They went back to school.

That’s what’s great about it because in Europe if you’re 35 and you want to go to medical school, that’s a non-starter. You’re complete albatross versus one of the people that we were renting from originally. She was an artist who in her late 30s, went to medical school and became one of the first, Da Vinci robotic surgery operators in OB-GYN here in Chicago. It’s a crazy story. That is not happening in Europe. Your path is set. You go from high school to college. It’s a different world.

When I describe you to friends or people that we know mutually, whether, in the industry or academics, you’re always the example for me of unbelievable passion, energy, and unrelenting work ethic. I’ve looked at my own drive and work ethic and sometimes thought maybe I should rev it down. I’m always like, “Kris works 5 times harder and is doing 10 times as many things.” You always make me feel better about working a lot. Where did you get that work ethic? Is that something that you saw in your parents?

I appreciate the compliment, but I vividly remember being lectured to my parents pretty much as long as I lived with them. I need to finish the tasks that I started. I need to be focused, do this, and that. In the summary you gave me, it’s impossible to do those things, and put up that vibe if you’re not doing something that you’re interested in. You can only force yourself so much. There’s only so much pressure on you, whether it’s from peers, friends, parents, or spouses. That can only go so far, then you’re going to burn out, crash, and burn.

You find a niche for yourself. That’s why it’s important to let these kids of mine and everybody else’s to find something that we’re interested in and not necessarily steer them against their passions because it’s going to be counterproductive. They’re going to revert back to whatever they wanted to do eventually once they get out of your sphere of influence. Having good habits is something you can do for them. For me, I enjoy doing what I do. I’ve been given opportunities to explore many fields. I found support for that exploration. We’re in healthcare.

The bottom line is we’re potentially helping other people feel better and be healthier. For me, that’s a pretty big motivator. If you keep that as the carrot on the stick that you’re here then there’s another person that’s benefiting from your work. Even if it’s just one person. We don’t need to save 100 million people. If we help one person and you see that person, which is something that’s unique to a physician, you see somebody you’ve helped, you could have a bad week or had plenty of meetings that didn’t go your way or whatever.

You have a clinic. You see a patient that you’ve helped even 1, 2, or 5 years ago come and see you. You are like, “Here, they come back.” They’re like, “I’m not seeing you for my back. I’m seeing you for my neck. My back’s doing great.” We are very fortunate to have that as a reference point. You don’t need to be a physician to be able to be part of that.

The guys in the OR, the companies, the nurses, anesthesiologists, and anybody that helps is part of that process. It’s like flying a plane. There’s the pilot, but you know, it’s hard to get off the ground if you don’t have the control tower, the guy putting the fuel in the tank, and somebody’s wiping that windshield. It’s a big-time team effort and it’s true in healthcare. It’s such an intensive effort on everybody’s part.

It’s a good segue because there is a lot of coordination that happens in your line of work as a spine surgeon and orthopedist. At what stage in your spine practice did you start to see some of these needs that became the passion and the drive to found HoloSurgical? What’s the unmet need you’re solving?

I was always interested in technology. I was always and still fascinated by people that have created something that other people use. Not obviously in the spine, but whether it’s the Da Vinci robot or something that’s a consumer-like Tesla car. It fascinates me that there’s one person or a group of people who put it together. Over a long enough period of time, they had a product that was beneficial for all. Everybody benefit and that’s amazing. That happens around us. You’re sitting in the heart of it in Silicon Valley. It happens there daily and some of those products are more useful than others, but they’re made by people.

I was always fascinated by healthcare. I also recognize that we’re antiquated as a sector of the economy. The thing we do, we’ve been doing some of them for many years. If you take the most common spinal procedure, laminectomy where we remove bone to take pressure off the nerves, those instruments were invented in Germany 1920s and they look the same. Maybe the metals and machines are a little bit better, but the principle hasn’t changed. On one hand, you can say that comforting that we found something that works for that long. On the other hand, there has to be a better way of doing that.

Failing multiple times is a sign of trying. Just continue doing them and they will work out eventually. Click To Tweet

I’ve always been interested in hooking up with people that have been very innovative and being a resident of the Cleveland Clinic. There was a decent amount of opportunity because that was a place in the middle of Ohio where you had a lot of brilliant people that were doing quite revolutionary stuff in the institution. I was in a lucky place from that perspective. They were doing a lot of trials. I was resident orthopedics and a lot of companies would come and speak with the physicians there and then try to work with them.

Those physicians were the people that got younger generations involved. That was a good ecosystem. I’ve had multiple interactions early on in my professional career that were supported not only by the institution. There was a pathway to get some support for your ideas, but also by strong mentorship on the doctor side or physician side and on the industry side.

There was a nice ecosystem there. Like most people that tried to innovate, you fail multiple times. I’ve certainly had, have, continue having, and hopefully in the future will have failures. If you’re failing, you’re trying. Some of these things will work out as they have for me. Some of those things work out years after you’ve thought of them and put them on paper, submitted some IP, or even did some basic work and did a prototype. For whatever reason, it may not have worked out then. In healthcare, there are a lot of reasons like, “The regulatory environment may not be favorable. The money may have dried up. The materials were not ready,” or whatever is working is too popular.

I’ve experienced that as well. I’ve always been interested in trying to incorporate software and some automation into healthcare. That’s a big thing to put on your plate coming from a medical background where there’s no coding class in medical school. That’s problem number one for me. How do I get into the software? I always want it to, but I wouldn’t even know where to meet engineers, then your budget. It’s not like I was carrying a treasure chest of cash that could hire people and start throwing up on it, which is another problem a lot of people have.

You have a good idea. You want to be in the space. You see there are a lot of things you can do. I guarantee there are people that are reading, whether they’re physicians or people in health, that are like, “Why are we doing it this way?” Most healthcare is like, “Why are we doing this? Why is my patient bringing me a CD with their images? This was 2020. They’re still bringing CDs. That must be the only industry for which they make CDs.” There’s plenty of things like that, and plenty of people saying like, “Why are you doing it that way?” That’s probably a good question. There are some explanations in terms of the regulatory environment and the need to prove some of these technologies, but it’s probably a lack of concentration and effort.

We’re behind other industries, which to me is a great opportunity. There are many opportunities in healthcare, whatever the discipline, and whatever level you’re looking at it from a rep or a nurse in the OR or somebody in the clinic. There are plenty of these opportunities. To answer your question, it is throwing a lot of ideas out there. Most of them not being amazing to put it.

A lot of them are being correctly criticized. The criticism was justified, but you learn from that. That’s what I had said about failing if you’re throwing your idea out there and you get a bunch of comments on it, that’s good. You want that. You want engagement. You want people to try to combat your thoughts and see how you can defend yourself.

How do you handle the criticism?

By paying attention to it.

Is there any particular criticism that is hardest to deal with?

It’s hard to deal with criticism. There are some of these things you’re working on for years. You show it to somebody and the guy looks at it for three minutes and says, “This is wasting my time.” I spent six years on this and you spend three minutes. I didn’t even have a chance to tell you why we’re doing something or we don’t even learn the whole story. Right away, I turned you off. You have to have thick skin, but maybe that person that I described is telling you something that you can use. That’s how I look at it.

I’m like, “This guy totally blew me off. I thought we had something amazing. We spent a lot of time on it. A lot of people spend a lot of time on it. We raised some money around this and we’ve been at it. Within three minutes, we didn’t capture this individual’s attention. Maybe we should change our approach. A lot of times, that’s that. Maybe your presentation is not great. Maybe you don’t know what that other party wants from the product you have designed while that product offers that service. A lot of these things can be easily adjusted. Some of the things you may need to modify in a not heroic effort, but may require some work. Sometimes the reality is there was no fit. Sometimes, it’s great engineering, but nobody needs it or maybe over-engineered for the problem you’re trying to solve.

When you take the product or service to the end-user that you imagine. You communicate what you believe is the need, and then say, “This is my solution.” That particular criticism is pretty harsh. The target user that you’re trying to solve problems for says, “I don’t think about it like this.” To me, that’s the feedback that is so valuable because you could keep going for a long time if there’s not a problem or the problem isn’t appreciated the way you see it.

UMN Kris Siemionow | Robotic Surgery
Robotic Surgery: It’s difficult to raise meaningful money, hire great people, and have everybody interested right off the bat. If you’ve had a huge exit and everybody’s heard about you, those problems go away to some degree.


Where I see entrepreneurs and where I’ve made missteps is the value proposition made sense for the target user, but the investor didn’t get it, or the acquirer didn’t get it. There’s a risk when entrepreneurs have a solution. It’s in a stage where you could get a product-market fit. If you could get to the market, maybe capital is your hurdle or some regulatory clearance. In the process of raising capital, not everyone has buckets of money sitting around and all of a sudden, make it a fundable idea and morph into something else. I would be interested if you had any experience with that.

It’s difficult in general to raise meaningful money, hire great people, and have everybody interested right off the bat. If you’ve had a huge exit, everybody’s heard about you. Those problems do go away to some degree, but then you’re facing some other challenges. Back to your point, one thing that I would do differently now, having the experience that I do, because we’ve been at this for quite some time, I would try to get in front of those groups or stakeholders as early as possible.

Provided you have the discipline to see it through because you will hear a lot of discouraging things. Get in front of the surgeon, if that’s the end-user as soon as possible. Let’s assume it is in this example. If you don’t have a surgeon, which makes it easier to some degree to put yourself as a customer, but then you become super biased, then you’re drinking your own Kool-Aid. Some of that objectivity disappears.

You have friends that are doing what you are doing or colleagues that are potential customers for you that will listen, and then probably spend some time with you. Those are valuable resources. The quicker you get in front of them, the better. We’re in this COVID situation and meetings are curtailed, but you go to an orthopedic, spine, or industry meeting and there are plenty of people that are in the finance or VC world that go there for that reason to meet other people. It’s a low stress. You’re not asking for any money.

You’re saying, “This is what we’re working on. Mind if I keep you updated?” and then find out what it is that those people need from you. You’ll hear some things. You’ll hear, “We only write a check if you’ve had FDA and $10 million in revenue.” That’s good for you to know because what’s the point of trying to get excited about getting in front of that guy until you do that?

The finance people are always pretty good at giving feedback on what it is that they need to see to get them excited and that lets you gain that perspective. The reality is the early money you’ll probably have to come in from you. If you surround yourself with a good talent for whatever you’re developing, then the chance of collectively coming up with a solution that gets people excited is a lot easier, especially if the perspectives are different.

I’m coming from surgery. We have guys that are coming from software that has never been in the operating room. Talk about a total divide between our experiences. They’re coming up with solutions that seem like they’ve spent some time in the operating room, but it becomes natural for them to think of problems in such a way. You try to get in front of people as soon as possible.

They say in the consumer world, “You want to get a minimum viable product, get it out there, and try to get people to criticize it.” We’re in healthcare. We can get a minimum viable product, get in the operating room, and start operating with something that develop three weeks ago. That’s a non-starter for us just because that’s not how it works. You would need FDA, tests, animals, this, and that. That doesn’t mean that you don’t have a concept, a napkin, a solid bones model, or some prototype you can show to people and see how they react. Are you totally insane or insane enough where it may work? It’s a very subtle difference between those two ends of the spectrum.

If you go back to the problem that you were trying to solve originally, what was the unmet need you were trying to solve with HOLO and were you insane?

History will judge. You do have to be very interested and passionate about something, which to some from the outside, may look like a bit of insanity. You need to be focused on and see that what you’re doing is important at least to you, then poison other people around you with that concept, and get them involved. What we were initially trying to solve was a problem of visualization. As a digital surgery company, we’re taking data and converting it into information.

The initial problem we were trying to solve is to show the surgeon the internal anatomy without making a cut on the skin. It’s not a novel concept. This has been reported in the ‘40s, ‘50s, and ‘60s. If you read sci-fi books from the s’60s and ‘70s, X-ray vision is a common concept. You put some pair of glasses on, look inside the body, and see what you want to see. That’s revolutionary about that idea. There’s nothing questionable about the need for that. People want that. People want to see inside without making a cut. Patients don’t want to be cut on so the doctor can see what’s going on. What do we do in surgery? We open things up and look.

The concept is not novel. The need is there both on the surgeon and the patient side. It’s on the surgeon’s side because if you see everything that de-risks the surgery for you, it makes it easier for you. I didn’t invent the concept. By any means, execution is where you can stand apart from other parties. Here’s when things get a little bit interesting.

We realized a couple of things. You need to be able to teach computer anatomy in order to be able to realize your original goal, which we did not have back then. We wanted to display the anatomy, take it, let’s say for simplicity’s sake, out of a CT scan, and use some software to process that imaging, throw it there, and match it with the patient’s anatomy, which by itself is a big challenge. in order for that to happen, several other things would have that happen, which are totally independent of us. When it comes to innovation, a lot of it is timing driven.

Focus on the important things, then poison other people around you with that concept and get them involved. Click To Tweet

The famous thing is Bill Gates was born three years later and you wouldn’t be where he is now. He would be a smart guy doing something else. Similarly, with us, we needed a lot of processing power in the form of graphics cards, graphical processing units, GPUs, and convolutional neural networks, which is one of the things that run machine learning outside of academic institutions. Prior to 2013, it would be difficult to process an image using these novel neural networks, which learn differently than what was the standard up until then. We are fortunate that other people developed tools that we could use to do what we do now and no one else was doing it.

The competition for that solution was non-existent since. The tools became available one year before it was before we started working on it pretty hard. Essentially what we were able to do as a result of innovation that was independent of us, was to create models where we teach computer anatomy. You feed a CT scan of your body and the computer knows the difference between a piece of bone, a nerve, or a vessel.

Within that bone, it can label it like you open up an anatomical Atlas. The computer knows the difference between a spinous process, vertebral body, aorta, or lumbar nerve. That’s what we start off with because I’m a spine surgeon, so we were doing a lot of spine anatomy. That technology is applicable across essentially everything else.

We’ve gone as far with it as teaching the computer lobes of the liver so that you can say if it’s in the caudate lobe or whatever. Where’s the tumor and what do you do with it? Once you teach Computer Anatomy, a lot of things become available to you. This is something we were not thinking of at the time, but it’s very interesting about innovation.

We taught Computer Anatomy, which allowed us to realize our initial goal to control that anatomy and be able to project it. All of a sudden you realize, “We taught  Computer Anatomy. It’s like teaching a human Surgeon Anatomy.” Once you teach Computer Anatomy, you can teach the computer how to plan surgery and the ideal placement of implants, whether they’re stents and vessels or screws or implants in the spine.

Since we are taught in anatomy, you can teach the computer to not do something, “Do not hit a nerve ever.” The computer knows anatomy. It knows what a nerve is. You can teach the computer not to hit a nerve. If you have a robotics platform, a popular topic, the robot does not know the difference between a piece of bone, nerve, or vessel. It does what the surgeon programs lit to do.

it’s essentially a surgeon programming the robot, but it doesn’t meet the definition of a robot, which should be autonomous. It doesn’t know the difference between the spinous process, pedicle, L4, L5, S1, a lesser or greater tuberosity in the shoulder, or some bone to deliver. If a computer knows that, it can avoid problems or at least signal you have an upcoming problem, “Doctor, this is a nerve because I recognize it as a nerve. Would you like me to continue on this path? Yes or no.”

It’s always a surgeon’s decision. Surgeon’s in the driver’s chair. They have to sign off on the plan. This creates an opportunity, an unmet need to prevent problems in the operating room, to prevent us, humans, from having a bad day, hitting a nerve or a blood vessel when we were not supposed to, or having that much better of an outcome by being able to pick the right type of an implant for that patient, whether it’s personalized in size and shape, but also whether it’s personalized because it meets the density needs of that person.

The computer knows anatomy. It can tell you what the bone density is in a particular part of the spine. It can create an FEA model, the Finite Element Analysis model, to see what stresses and strains theoretically occur in that part of the spine. All those things happen only because you taught computer anatomy. I’m oversimplifying some of the things that need to happen for it to work efficiently, but that’s a general concept. Our initial goal was to show you inside your body. In order for us to do that, we had to do a number of other things that we were not thinking about initially. I created so many opportunities, which quite frankly are much more valuable in my perspective, than our original ID.

The ability to see into this anatomy and have the precision of something like a CT scan is highly valuable and interesting, to avoid having to look up at a monitor and keep your view on the surgical site. I always thought that made a ton of sense, particularly as more computer-guided surgeries enter the OR, more screens show up. It’s just another distraction.

Sometimes I sympathize with the surgeon. How many people are talking to him or her? How many things are beeping? How many screens to look at? If I understand it correctly, you couldn’t do that important thing, unless you had enough computational power to process all these graphics that the GPU constraint. I didn’t realize the neural network component of this. What happened in technology that enabled the neural network that you’re building now with HoloSurgical?

It’s a confluence of things. You’re sitting at the epicenter of it where they’re smart academics. It’s Stanford, but also a lot of finance companies like Intel that saw the value of putting more money and making these processing units and the software that runs them not much more robust. It’s this whole concept from the 1950s of trying to recreate how the human mind perceives an image and translates it into a computer language. A lot of people don’t know how these things work. You can train this model as you would train yourself. We have a whole team at the company who is like the data marketing department.

It would be interesting for people to understand how training the computer work. Essentially you take the images and hand mark them, which is a tedious and painful task. You can imagine every slice on a CT scan. Some CT scans have 400 slices. Every one of those slices is to be labeled by a human being initially because that’s how the training process works. It’s something that’s supermundane. At the same time, you need somebody that’s super experienced doing it. You can’t just have somebody down the street, a bunch of high school kids marking things up. You can do that for some of the other algorithms. This is a stop sign, yes or no. When it comes to say, “Does the nerve stop here? What starts there?”

UMN Kris Siemionow | Robotic Surgery
Robotic Surgery: Robotic surgery aims to reduce and eliminate the error. But as long as there are human in the loop, errors are always bound to happen.


That was one of the challenges you need to solve. Once you get enough of this data, that’s the question, “How much is enough?” That’s not an easy answer. You feed it into this network, which is essentially a computer program that does a bunch of calculations on these GPUs. You verify your results. The results may look great on the computer. You may have great scores that reflect high accuracy, but then you give it to a human. You’re like, “This is nothing that we expect.” You go back and start tweaking the way these calculations are made, where you put more weight and emphasis on which parts of the training and what’s more important.

You have to prepare your data accordingly. It’s not an overnight process. The concept of the network is there. There are a lot of networks available to people that put a company together. Let’s say you get 1 or 2 software engineers and start playing with these things. There are plenty of open source networks that you can try to reconvert. In order to be super-efficient and practical, you will need to spend a lot of your own resources and time to design something that’s pretty custom.

I want to make sure I understand this correctly. When Google was doing Google Maps, they would go out and send these cars. They were trying to get videos to map the United States and then eventually the world. That was a popular product. Apple did the same thing. What Tesla then could do once a mapping environment existed, you can go into other cars. If I back my car up, it senses something and stops me. Something like autonomous driving. Is that a neural network? They’re saying, “I’m this many meters away from the car in front of me,” or is that something different?

It should be the output of the neural network. That’s what the neuro-level work was used to train the car to do. Technically, you can get into a neural network and say that it can happen in real-time, but first of all, you should be able to train to do the things I mentioned. In our case, it’s marking up and training the network. If you’re giving an example of Tesla.

Tesla would have to be able to recognize the difference between a stop sign and a go sign. The neural network was used to do that and then you developed an algorithm inside the computer of the car. Information is fed from the multiple sensors and cameras that the car has into that computer and ran through that algorithm. The output of that is, “Drive. You turn left. You slowed down.”

Could HoloSurgical be the neural network and algorithm that makes robotic surgery as it is now truly autonomous? What’s the guidance of a doctor?

For disclaimer purposes, we always need to say that the doctor is always in charge of it. You do want a human in the loop, despite the fact that you probably don’t want a human for most of these tasks to be executed. The platform is a software platform. If you use a robotic arm by OEM. There’s a decent amount of companies that are simply making robotic arms. What you are getting, if you ordered this thing online in your office, is a robotic arm that looks super futuristic and has a bunch of mechanical parts, electric motors, and wires. There’s your robot. That’s the point. Why isn’t it doing anything?

What drives the robot and the business is in the software. How do you make that robot intelligent? You have to teach it things. You can teach it to vacuum your room. You use an artificial intelligence neural network to teach it like the difference between a long-haired carpet and a short-haired carpet and have different modes for that, mop the room, and all those things, and execute on that.

Similarly here, we have a robotic arm, which is a mechanical device with a bunch of engines and cords. It doesn’t know what to do. We put it over the patient in the operating room and we hook it up to the bed. While it still doesn’t know what to do, it needs to be fed something. It needs to be fed an operative plan and something to execute. The only way this robot will know how to execute the plan is if it knows where it is in space, in relationship to its target, what s the target, why is it there and all those things that need to be answered. That’s where the algorithm comes in.

We would scan the patient in the operating room with a scanner. We would generate a DICOM file, which has a bunch of data. We need to convert that data into information because the computer does not know anything about that patient. What our algorithm does is highlights all the different things on that image, whether it’s the spine, nerves, vessels, or muscles then make a bunch of suggestions that the doctor has to select, “I want to do the surgery at L4 and L5.” You push the button and the arm steers itself based on that if you’re using an arm, but you don’t have to. You can just make those suggestions to the doctor.

If you think of some of the surgical planning software that exists, you have a CT scan and the software that allows you to simulate what a surgery could do. You’re able to look at the anatomy and then plan what you want to accomplish maybe remove this or put something in a certain space. With surgical planning, is it more taking the place of the algorithm? Is the surgeon in a sense, writing the algorithm?

In your example, where you download something, get a ruler, plan, click, and move some implants in a space on your computer by clicking a mouse away, you’re the neural network. You’re the computer. You’re the thing that has been taught over years of medical school and residency to do this. That’s what a surgeon’s job has been in the 21st century. That’s what you were taught to do, and that’s what you do. There’s no reason in 2020 that you have to do the clicking and recognize. We have proven that you can’t. All you have to do is sign off on the plan.

That plan gets generated for you so you have more time to see more patients, do more surgeries, and relax with your family because planning takes time. That’s what it is. You have to go somewhere, load the data, look at it, and think about it. That takes time. If you do ten procedures a week, that’s 10 to 15 minutes per patient. That turns into hours that you’re spending doing a task that can be performed by a computer.

If you surround yourself with good talent for whatever you’re developing, it is easier to come up with a solution that gets people excited. Click To Tweet

The neural network has the networks of facts that the more it’s used, the more information that goes into it. Whereas the surgeon, even when someone is as well-trained as you, your plan is going to be influenced by what you were taught, read at journals, or heard presented at meetings with a heavy bias towards what you’ve done before that worked.

I’m not benefiting from other people’s experiences at all. We all know there are amazing doctors up there that are busy that see a lot of pathologies and are doing a great job. There’s no way for me to benefit from that. I can watch all their videos, YouTube, this, and that, but that’s not the same thing as having the computer have that information over thousands of cases and doctors.

I don’t want to get into regulatory things, but having an unsupervised network, you’re feeding data, and continue churning out results, has some limitations from a regular perspective. You don’t want this thing to not be validated at some time points. We want to feed it a bunch of information and have some smart people that know what they’re doing, tested stress, tested, validated, throw it some crazy examples, see what spits out, perhaps even do some type of bench testing, and then release it.

You can still release these things every week, 6 weeks, 6 months, or 6 years, whatever your cycle is. My point is you have something that’s locked, that’s been tested, stress-tested, and you’re happy with. That’s what you deploy into the field. All the while you’re collecting data, analyzing it, and testing it against your current validated network.

Open source networks for military drones would be irresponsible because the stakes are so high. It’s the same stakes in medicine and surgery. All those surgical have built and are developing the neural network and the algorithms that are going to enable this type of surgery.

That’s why it’s easy to be high-speed all the time. That’s your end goal. It’s also easy to get other people involved. This is not something that’s only unique to us being in this space. There’s plenty of room in healthcare to be doing these things. It’s easy to get up in the morning and convince people to join you on this quest. The reality is even if it fails from a business perspective, you still were involved in something interesting and exciting.

At the same time, I do feel that there’s such a need that you’re limiting that business failure risk, essentially to your risk of execution and your risk of the regulatory environment. While we were talking about some futuristic stuff, you obviously want to be within the constraints of what is allowed by the regulator presently.

The regulators protect patients and that’s an important role. I read a statistic, it was about 90,000 patients in the United States alone have died as a result of the Coronavirus. The point of this statistic was that doesn’t put it in the top three. It’s maybe incorrect and you know better. Medical error is number three of the leading cause of death since the Coronavirus pandemic in just the United States.

It seems like medical error could be a number of things prescribing the wrong drug, or not knowing an allergy, but surgeons as skilled, trained, and hardworking as they are, no one’s 100%. If you have a technology that can reduce medical error, add precision to the technique, enable the surgeon to spend his or her time with the patient and less clicking on things that has to have a big impact on not only improving patient lives but maybe saving them.

What you’re highlighting is important. There are a lot of opportunities to reduce the error. Our goal is to eliminate errors. As long as there are humans in the loop and the system is structured the way it is, there is always going to be room for error to happen. Any one of the technologies along the spectrum of care from the moment the patient walks into the office all the way to the time done with care, I don’t even want to say discharged because once you’re discharged, you’re still under care and errors can happen. There’s so much opportunity for all of us to contribute to reducing that. It has gotten easier for even small teams, 1, 2, or 3 people, to be able to come up with a solution that’s widely adopted. That’s not super complex, but has an impact on cutting down on error.

The access to computational power is essentially free. Everybody’s got a smartphone. You have an idea. You don’t need to know how to code anything. You hire coders. You can do that online without ever meeting them. You just tell them what it is that you want. For a very reasonable budget, you can develop a solution, stress test it, show it to people and see if it’s got legs. It’s a super exciting time because the need is so big. There are plenty of opportunities.

It’s not like, if you’re doing artificial intelligence in healthcare, you can apply it to essentially everything. It seems like a very generic term and a very big buzzword, but that’s reality. There’s so much opportunity that it would be silly if somebody’s truly interested that they don’t try it. It’s like one of those mottos, “The biggest risk is not taking a risk.” That’s true. If you have an idea, what’s the worst thing that can happen? Maybe you’ll take a financial hit. You control how much you’re going to spend.

Especially if the drive is coming from what you’re interested in and your passion. You make a change and less about getting an A-plus on a piece of paper. Before we move to the vault, one of the things I learned sometimes the hard way as an entrepreneur is when I’ve been focused on healthcare innovation, where the patients who I’ve had in mind from day one. That’s my passion and drive. What I’ve learned in things that have worked better than others is when the healthcare provider, in this case, a surgeon is the person that’s going to take that technology and deliver that value.

UMN Kris Siemionow | Robotic Surgery
Robotic Surgery: Even when talking about futuristic stuff, always stay within the constraints of regulation.


If there is also an unmet need for the surgeon, you’re helping two people. There’s going to be a higher incentive for the surgeon to adopt the technology because it’s going to make his or her life better. You said something that a lot of people can relate to. You’re busy. You have a family. You have other hobbies and interests outside of surgery. If HoloSurgical can do everything that you’re saying, think of the time you’re saving for the surgeon. Not only can you treat and help more patients, but you can also go out and have a great life.

One of the headlines of the Coronavirus crisis has been the “brave healthcare workers.” It was kind of you to recognize all the different specialties. As the world and certainly in the US, we need more and more people to choose their careers as healthcare providers. As technologies like HoloSurgical make that job safer, easier, more rewarding, and less mundane on the things that computers and machines can solve, all of that is going to contribute to better care and care that we can continue to scale that maybe now we have some limitations.

One of the things we want to do is cut down the amount of nonsense that happens in the office before the patient gets into the operating room. This is no cut on the insurance companies, but they do certainly take up a lot of doctor’s time, and back-office time. In order to get the surgeries approved, they require a lot of documentation. It’s getting only harder and harder.

Insurance companies are smart. They’re using artificial intelligence to tell people what their opinion is about how to take care of the patient. There was always a famous contrast between how a physician and insurance company thinks about it. What we’re doing is using imaging essentially to be able to help the doctor demonstrate the value of that treatment plan.

It’s not always surgical. If you go to the surgeon’s office, it’s not like you’re always getting surgery. You are getting a lot of other things like physical therapy, may be an injection, or an off-work note. The best way to support that is an objective way. We all know that there’s plenty of data floating around in physician’s offices. Everybody knows your height, weight, left leg pain, 63 years old, and diabetes.

All those things are available. There’s nothing novel about that. It’s difficult to take that demographic information, even that historical information to prove your point, although that is a necessary component. Objectively, you can tell from imaging what is going on with that patient. You have to be able to quantify the image.

Once you’ve taught Computer Anatomy, you can make some predictions and correlations. You can tell which patient will benefit most from what type of treatment. You can generally do that in a form of a report. You can do that rapidly using a computer so that your back office, your PA, and you as a physician don’t need to spend time proving your case to the insurance company. You have an objective way of saying, “A person with a spinal canal, this diameter, or joint space narrowing in the left knee of this diameter has been shown over thousands of cases to do well with this treatment.” That’s something I’m very excited about because a lot of clinics are very painful for the physicians, mostly because of bureaucracy, not because of patients.

If all I had to do is walk into the room, talk to a person, and be like, “I’m going to send you for physical therapy. You would benefit from a surgical procedure,” that would be easy and enjoyable, but no. You have to dictate. You have to fight for it. You have to code and bill for it. There’s no reason why people are doing that. That doesn’t all those things can be done by computers. A lot of other individuals are thinking about solutions for that, too. It will only get easier and better.

It’s an exciting and hopeful time. Let’s go to the vault. The first question is over 2021, I know you read a lot, what’s one book that not only did change the way you think about maybe your work as an entrepreneur or as a surgeon but overall had an influence on your life you’re still thinking about it?

I love reading. For me, it’s been Lifespan. David Sinclair is a PhD. He is a pretty well-known guy. He talks about why we age and why we don’t have to. That blew my mind. I’m a physician. I like to consider myself somewhat of a scientist, but I was embarrassed that I have not explored the concepts that he discusses sooner in my life. A lot of these things are things you can do to live a better life, which is fun. It’s not all hypothetical. The research he talks about that’s going on and many brilliant people that are working on that’s on the research side. There are plenty of companies and money in it.

It’s an exciting environment that I want to be involved in once I free up some time on my plate. I’m glad there are a lot of smart people working on it. I’m positive, especially when we’re dealing with all this misery from Coronavirus, all these people suffering. It’s horrible. To have that positive perspective of us being able to, not in too distant of a future, implement some strategy, pharmaceuticals, or other gene types of therapies.

That will be safe and help us live not only longer lives but healthy lives. Who wants to live a long miserable life? We want to live long healthy lives, and that’s the premise of it. The book makes you think about how you live life and there’s still a limit to how long you got. There’s a level of a philosophical component to it, which I also enjoyed. That’s the book I’ve been promoting to everybody.

Thinking back over your life and career, aside from your parents and their accomplishments, who’s the one person that took an interest in you could have been when you were younger, maybe early in your clinical career, but that saw some potential in you and gave you the encouragement for you to continue and make all this progress?

Learning how to break complex things down is a sign of fully understanding them. This allows you to make them better for yourself and for others. Click To Tweet

If you don’t mind, I’ll answer the question my way, because you asked like, “Who’s the one person?” All of us have different stages. There are three stages through a surgical procedure the opening, you’re in there doing what you need to do, and there was closing. All of them are equally important. If you screw up 1, 2, or 3, the whole surgery goes to hell. Similarly here. You’re a kid, teen, college, a young professional, or a middle-career type of person. For me, I’ve always had somebody that was not a parent that saw something for each one of those stages.

I probably asymmetrically benefited more than that individual. I don’t want to say I feel like a leech on this person, but I tried to be a sponge. There was an orbit and things fell off if you were within that orbit and part of that entourage. When I was a resident, there was a particular surgeon who was entrepreneurial, was always open to involving other individuals, not just myself, collaborative, and had a good business sense, but at the same time did not screw these little people like me over, showed me that you can be successful and incorporate other individuals. Now that I’m in this stage, and there have been different types of a mentor, for many reasons, some of them are as simple as geographic.

You’ve moved to a different place. You met somebody new. You have a different interest than you had before. It’s not that it’s just one person I can say put up there. For each stage of your life, there’s somebody, whether it’s a hockey coach when I was in high school who has nothing to do with medicine and entrepreneurship. This guy happened to be an entrepreneur now that I think about it. I was there for a different reason. I was there to learn hockey and be encouraged.

Realizing that such people exist is so valuable because, first of all, you know you’re going to be okay because there’s always somebody out there that feels like that’s their responsibility or whatever mission to help others. It puts the responsibility back on you because you are busy. You’ve got stuff going on. You got your own family and kids. That’s all of these other people.

They weren’t there because they couldn’t sleep at home at night and had to get out. They got other stuff going on, but somehow they found time for you. Maybe you should find some time for other people, too. It doesn’t take that much, but include them and see what happens. It’s such a synergy. It’s not like it was a unilateral one-way street.

I’m excited about finding people that are interesting, no matter what their age is. Sometimes they have ridiculous ideas. They’re like, “I wish I was like that.” Unfortunately, because we have a residency program, so we have a lot of young, smart people that I’m questioning myself whether, “I was like this. These people are blowing me out of the water,” and they’re half my age. That question is exciting because it creates a lot of positive buzzes.

I like the concept, the beginning, the doing what you need to do, and then the close.

Lester Borden, a famous orthopedic surgeon at the Cleveland Clinic, did not use any cutting guys. There are tons of legends about this guy. He did a perfect knee replacement in an efficient manner. He didn’t make any mistakes. He is a legend amongst orthopedic surgeons. He had his own fellowship, but he would teach these simple concepts. You can make fun of orthopedic surgeons that they’re pretty simple people. There’s probably some truth to that.

When I was growing up in Poland, there was a famous Polish soccer coach who won the bronze or the silver medal in the World Cup in the back of the ‘70s when they had a great team. He would be the guy that teaches you simple concepts like, “The ball is round. There are two goals. He who scores wins. That’s the game.

Similarly here, you can break everything down as a simple concept. Something as complex as surgery, all of a sudden made you say, “If I work as I’m learning, as I’m a resident,” if you’re a rep, the guy is opening. “What am I doing when the guy is opening? The guy’s doing what he’s doing. What’s my role then? The guy is closing, what do I do then?” Breaking these things down is super beneficial because once we break complex ideas down, first of all, that means we understand them. Once we understand them, we can make them better for ourselves or for others.

You need to know what time it is to know where to focus. In building a HoloSurgical, there are consultants, third parties, and vendors. Was there one problem that was difficult and there was a particular service provider that does a great job?

The answer to that question depends on your personal background. If you’re not a physician, trying to develop for physicians, you’re probably seeking out physician consultants as a high priority for you because you have a lot of questions. You don’t know a lot of things you’ve been doing. We didn’t have that. Our unique problem was regulatory. It’s this a big animal, that’s important. There’s so much emphasis placed on a company in a regulatory pathway, clear ends, the strategy around them, this and that. For us, it’s having help in that domain. That’s a big change for us.

You can hire smart people. I’m not saying you need a consultant if you’re a startup. Essentially in R&D, we have twenty-some employees, but a few are our engineers who are well-put-together people, but they are not experts on regulation. We needed to recognize that’s our problem. First of all, you need to know what your problem is. Second of all, you know when to tackle it. If you get a regulatory consultant in 2015, and now it’s 2020, is that too soon, or is not soon an offer?

UMN Kris Siemionow | Robotic Surgery
Lifespan: Why We Age―and Why We Don’t Have To

When you think about a concept, it’s important that you know your regulatory pathway and who’s going to pay for it. Maybe you will develop a great product that doesn’t have funding, and doesn’t have clearance to answer the market. All those things are important. My eyes roll to the back of my head when I think of regulatory work. At the same time, I recognize that it’s one of the most important things we do for the right reasons. There is a lot of this Q&A stuff. You’re essentially forced to do the right thing and that’s important.

Is there one regulatory consultant in particular that you think does a nice job?

I don’t want to maybe give a particular company name if that’s what you’re interested in. We were happy with who you use, but that’s not to say that we haven’t had a good experience with other people.

If you think of all the great software tools that all companies have access to, what’s one software tool at HoloSurgical that you found, adapted quickly, and you can’t imagine working without it?

A very popular thing is Python, a computer language that helps with machine learning or artificial intelligence. A lot of that is as based on it. The good news is other people are developing it for you. This is a community thing that is being developed as you sleep or work. You benefit from a lot of synergy and brain power. Your role is to essentially have people you’re working with utilize those tools that are widely available. Most of them are essentially free to answer those questions that you’re posing.

That’s on the computer. There are multiple providers. The good thing about software is it’s been around for so long. There are so much R&D and capital in it that most of the tools are there. Your job is to put them together and maybe fine-tune them for your application. It is not to say that there’s no IP protection or opportunities of that nature. You can buy off-the-shelf software to create very unique things for which you have granted patents that allow you to be the only game in town.

Last question, what is your biggest unmet need as a business?

I’ll have multiple answers. It’s not one thing ever. First of all, we are living in very challenging times. In multiple businesses in healthcare, I have been appended by what’s going on. We need to know where we are and we need a little bit more time before we have more clarity on that. We are not immune to that. We are in the business of providing software tools to doctors and they’re not for free. Those doctors and hospitals have to feel comfortable enough to be able to allocate those resources. That’s one thing.

Other than that is finding enough people to help us grow as we enter the market. It’s easy relatively to find a couple of great people. As your organization grows, because you have this culture that you’re not trying to disrupt, you have a pressing need, whether, “We don’t have any salespeople. We need salespeople. How do we integrate salespeople with our R &D team?” There are multiple books that have been written about that in Silicon Valley, especially. Those are some of the challenges and needs that we will have to build a team that reflects who we were up until now and has the same passion and desire to bring our technology to the user.

Kris, thanks for going to the vault. His favorite book in the last year was Lifespan by David Sinclair. Multiple mentors throughout your life. The biggest unmet need at HoloSurgical is finding the right people. Regulatory is a very key service. Kris, I can’t thank you enough for all the time. Thanks for being on the show.

Thank you.

Important Links

Leave a Reply

Your email address will not be published. Required fields are marked *