In this episode, we chatted with Dr. Zsuzsanna Puspokis about how she co-founded Rhino3DMedical, a medical 3D printing-focused software based on Rhino 3D. Zsusanna demonstrated a few surgical examples in this interview.
About our guest:
Dr. Zsuzsanna Puspokis
Zsuzsanna received her PhD from the Swiss Federal Institute of Technology, Lausanne (EPFL, Switzerland) in 2016, where she completed her thesis at the internationally renowned Biomedical Imaging Group (BIG) headed by Prof. Unser.
Following her doctoral studies in Biomedical Imaging and before starting her own company, Zsuzsanna worked for a year at the Lausanne University Hospital (CHUV, Switzerland) on the Human Brain Project, a flagship project of the European Union. During her employment at the hospital, she further strengthened her industry and academic links and deepened her knowledge of MRI imaging, from data acquisition to processing and analysis, in the clinical context.
Since 2017, she is a co-founder and the COO of Mirrakoi SA. The company offers software solutions (Rhino3DMedical) for orthopedic and CMF surgical planning, facilitating the creation of patient-specific 3D printable anatomical models and surgical guides from medical CT/MRI scans (https://rhino3dmedical.com/). At Mirrakoi SA, Zsuzsanna leads the design and implementation of business operations. She is responsible for marketing and sales, including PR and customer relations. Zsuzsanna is well-connected within the European start-up, venture, academia, and entrepreneurship scenes where she has been active for several years and recognized by various awards and grants.
Hey there. This is Jenny Chen. I'm the founder of 3d Hills. Welcome to the lattice podcast, the official podcast for 3d Hills. This is where you will find fun, but in depth conversations with technological game changers, creative minds, entrepreneurs, rule breakers, and more focusing on how we can use 3d technologies like 3d printing and bio printing to reinvent healthcare and even Life Sciences. This podcast will also include ama, or ask me anything sessions, past Instagram Live interviews with influencers and other direct engagements with our tribe. Hi, everyone. Very good. Thank you, Susanna, for joining us today. And thanks for everybody who were tuning in for this episode of Instagram Live. Who we have today is Susannah Pushpa key. She is the co founder and chief operating officer for Rhino 3d Medical. I've been following you guys for quite a while. And so Susanna, you actually just graduated pretty recently in 2016, after you complete your PhD in a in a Swiss technology, Institute of Technology. Yeah. And actually your study was focusing on Biomedical Imaging. And after that, Suzanna went on to work at a hospital, I guess focusing on Biomedical Imaging for another year, where you are acquainted how to optimize data acquisition was CT MRI data. And then after that, in 2017, you founded Rhino 3d Medical, which is incredible startup that I'm having. I think there's a lot of potential and I want to give you some time to talk about that today. So Suzanna, how did you decide that? There's, there's quite a few questions I have for you is, why did you decide to start or co found Rhino 3d? And why using Rhino as the base for your, for your app? Alright, so thanks a lot for the introduction. Um, you said four years or five years ago, I graduated as a PhD. And actually, for me, it already seems like a very, very long time ago. So miracoli has been co founded in Yeah, in January 2018, by the story dates, really, really back during my PhD studies where I met the other co founders, Daniel, and Pablo, and people are working on different image representations and segmentation. And in particular, what was interesting is really to find a segmentation for of special specific anatomical, and then give the rights to the user to add it this kind of segmentation in a very easy and handy way. And after my PhD, that's true, I spent a year at a local hospital here in in Switzerland. And there I was working on a project of the European Union. And I was really gaining a lot of experience in in MRI image acquisition and also a little bit in in CT image processing. And during these times, you know, we had a lot of active discussions with my previous colleagues from the Swiss Federal Institute of Technology. And we really had the idea Okay, let's try to bring it to the industry but we were working on for so long, and try to see if it can bring benefits and added value to the people that we think it could be. And back then it was still unclear for us what we exactly wanted to achieve. We knew we loved 3d printing we knew we love the technology and interdisciplinary field of a little bit engineering informatics, mathematics and biology or the medical field by for still unclear what would be the exact goal. So at the beginning, we were still somehow finding our road and we built up a very strong strategic partnership with Dr. Sarah us. I think we originally chose rhinoceros because it's one of the most common new software for organic and freeform design. So if you think of solid modeling, this is something else this is really the organic freeform smooth modeling and this is somehow the word where we are coming from. And we gain experience, let's say in developing plugins and features with Rhino for Rhino and we established this partnership first with the European Office of of McNeil, which is in Barcelona. And they really supported us from the beginning. And then finally, I think we, when we came with the medical idea, it really became like a long term partnership because somehow Rhino is lacking this medical image, let's say they are pretty strong in architecture in product design in industrial design, but not for for medical imaging. So I think what we are trying to do is to build like a complete medical imaging platform with a little rhinoceros heart, reading that. So I before before meeting you, I didn't even know the existence of Rhino sorry, as a software because I'm not in. I'm not an engineer. I, the number of cat design software I know is a handful. But recently, I realized how much people love random story, actually. So and you've encountered a variety of software in you know, was cat feature. What do you think? One Rhino sorry, actually does butter? You mentioned freeform, which I'm not really sure what that concept means. If you can explain to us that'd be great. And to compare it to the other CAD design software, what is the advantage of using Rhino? I think our CAD design software is mostly used for like mechanical design, like designing machines where you have a solid volume, like you define a rectangle, for example, and then you do extrusions. solid modeling is like when you give a playdough to a kid. And then he really manages to create smooth shapes all around, that has not really like volumetric form, but something like like very smooth and very curvy, like really a piece of art, let's say. Cool. And I know that you bought a couple of examples today, for us to demonstrate the power behind why no 3d Medical software? Why don't you show us those? Yes, so I prepared two examples there. There'll be let's let's talk a little bit in general about the company and then I could show my models. What we try to do is to connect the medical imaging modalities like CT, and MRI imaging to your 3d printer as easily as you would print out your document from the computer with the regular printer. So this is the main idea and we are focusing on three main areas which is the CMF surgery, orthopedics and trauma. And my first example will be from the, from the CMF for the cell. And one of the most common frog fractures that you could have is the fracture of the orbital floor. And if I yeah, maybe I can come a little bit closer and add a light. So, here you can see the orbital floor which is the healthy side and this one is the slide where you can see the fracture down there. Yep. If it goes to through the camera, so, here the preparation for the surgery is pretty complex in the sense that you have to insert a plate and that will have to cover the whole fracture and you have to be very precise with that and it has to cover the entire hole. Otherwise the tissue behind it can can come over. So probably in terms of medical medical terms, he could explain the surgery surgical procedure better than me. But I think having a print in your hand which is very easy to prepare, like from the moment you obtain the daikon images or nifty images and the CT scan of the patient you can select the area that is interesting to you. And then you can directly send to your desktop printer and you have it just before the surgery you can already select which implant you want to use, how to prevent it, you can discuss with your colleague the entire procedure, not just based on like a stack of 2d images that are in your computer, but having really like a tactile information in in your hand can really help also we can talk about patient communication, yeah, and and all that. So, what we believe is that if you have these models, it can really help you know to to go to the operating room where all the questions are basically already answered. And such that the the time you spend in the operating room can be much shortened Which is okay in one hand, it's good because it saves costs, but it's also less procedure risk for the patient. And this is I think, what we are really trying to focus on. I think everybody's dream that if we just push a button, then we can print out a model. So on your website, it says, Rhino, 3d Medical is the fastest software to make that happen. Why is it so fast? How do you how do you guys make it so fast? So what I believe is that there are some other software that is really complex and very high level, the issue I see is that most of them are really targeting engineers. And they are focusing on on different kinds of applications. So if I would like to use analogy, I would say, for example, if you just want to crop an image, you wouldn't buy Photoshop, right? Yeah. So I think the key is really that we want to give a tailored solution to the surgeons, or to the medical professionals, such that they can use the desktop 3d printer with them, and then directly just focus on the workflow and the operations that they need to have, or that they will go through. And they don't need to learn years of becoming half an engineer to be able to operate such a software. And I think, I think in some sense, our main competition is not like other companies, I think we are giving a complimentary solution. What we are really facing as an enemy is like the not knowing of the technology, the lack of using 3d printing within hospitals. So I just read recently, a statistics that around 99% of the hospitals are not at all exploiting the technology. And there are many barriers that I see. And one of them is that if you want to invest now, as a hospital in in 3d, 3d printing, typically, the option you have is the very costly, you need to buy some huge industrial 3d printer, you have to hire a group of engineers. And then if you have something to print out, then you send your request to this department, and then they will process it, depending on their priority list. Or you engage a third party company where also no patient rights come in, you know, it's confidential data that you are, you're handling, so you can engage like a third party company, that will do it for you, and then ship it for you. And then maybe the printer arrives within a week. So this is now very encouraging, let's say to to jump into this 3d printing. And what we feel is that Oh, it also aligned with the perception of people with 3d printing. So ideally, imagine it as like sky rocketing, or something very, very geeky that, you know, it geeks are doing in their, in our backyard, between playing video games. So I think there's a very broad spectrum in between. and it's exactly where we try to position ourselves to really make this technology accessible for everyone who could benefit from it for from it. And I think the best of 3d printing, how the industry is evolving, it's really in this direction, you can buy very reasonable 3d printers for a couple of $1,000. And I think this is really has the quality, such that you can do your prints within your office. And then you can use it to do have a better understanding of of the anatomy of your patient. So one of the other applications I know you guys do is surgical guides, which I find very useful because you know, if anything, one step beyond just having an anatomical model at hand is like something you can actually use in the operation. Does your software make it easy to produce these guys? Because that requires you to know some kind of engineering, you have to have some kind of engineering background. For people like me, I'm a radiologist. I wonder, you know, can I use your software to design guides for my surgeons? say exactly, that's the second step that I just wanted to show you. So the first module it was printed with an SLA printer, this one I printed just with a filament based printer, but we can already see the vertebra as it's their cervical vertebras Yeah. And here, the procedure is pretty complex if you want to insert pedicle screws, so probably all the medical doctors, they know that we are talking about like four or five millimeter in terms of hate, and then you really have to be precise if you put the pedicle screws from the right angle, because otherwise the nerve root that is around will be damaged, and it can cause serious life lifelong problem for the patient. So typically, there are two ways to insert pedicle screws, there are minimum invasive surgeries, and there's really the open surgery with either the free hand technique or with surgical guides, as as we see, even though these minimally invasive surgeries they are taking place, the procedure is complex or risky, it's still the Open, open technique that is used. And I think it's very useful to be able to create a guide very quickly for the patient such that it really adapts to the anatomy of the vertebra. So today, it's really really 100% sure that the way you are drilling the hole and placing the screw, that's perfectly to the to the anatomy of your patient. So we ourselves really tried to make I can easy workflow for the doctors and this module we have developed together with our collaborators in Brazil and, and in London, in in the United Kingdom. So here, the idea was to define different guide models, to the cervical to the cervical spine. And then if you read our blog post, on our website, this is the place of publicity, then you can find actual use cases where surgeons were using our our guide techniques for for the thoracolumbar spine. So here you can really see for each vertebra, you can really find a guide or and then you can also adapt to the camera. Sorry, can you hold it closer to the camera sorry. So you can also like design the these base parts such that it really fits to that part and this, this you can also close such that, you know, when you place it there, then it will really does just hold this up. So it's more to show the different options. And we are really talking you know about clues that are like couple of millimeters thick. So you really have to be precise, and you have to do it fast. So we are we are working on right now that I don't say we have a magical solution yet. But let's say we are on the way. Well, first of all, I'm going to post your block lens. Your block link on the on after this video to answer is a question from Ryan Odom. as relevant he says how does the quality of medical imaging and then the various isn't a type of medical imaging, I guess CT versus MRI versus ultrasound affect the amount of geometry. So we do ultrasound, we have less think we have less collaboration. So if you have some images to send us then I would be happy to assess it together with you know, typically, we CT and MRI imaging, we were having pretty good results. Do you have the kind of quality control or validation process for that? Yes. So anyway, there's the medical doctor that is there, controlling the whole operation. So we are not like a decision support system. So it's, you know, if I give you the tools, you you still have to see, you know if it makes sense or not. Now we have clinical trials, great, I could say basically, you know, over all the continents around we have we have quite good collaboration in in Brazil, Colombia, in the US in in India, etc. in South Africa and with those collaborations now we are doing clinical trials and what we are trying to prove and go through is that really how how well these these guides fit and and are useful during the surgery and right now we are having very positive positive feedback. Yes. You know, sorry. Go ahead. Sorry. Go ahead. Yeah. So if you you compare it with software that is already for example, FDA certified for most of the cases that concern bone, like bones CT or bone on MRI, you don't really see difference. So I have another question is relevant, I think it will lead up to the next topic that we want to talk about. So Mackey's iakovos sorry if I pronounced your name wrong, but he asks, I use open source software like slicer 3d, and meshmixer, which I also do personally. How do you compare your software to the existing open source ones? Like Blender? I think a lot of people use Blender as well. How do you compare? I mean, I think usability i think is a huge thing that you mentioned, I think it's it's mainly usability that you want to get your results quickly. And I think you need to be a little bit geek to, to master blender and those tools, oh, I think you have to what we want is really something that is workflow based, and that is, you know, easy to use, and that is dedicated for this given task. So you mentioned that most of the applications, so those two cases you demonstrated are in orthopedics and CMF. Why focusing on these two applications? I think, first we started with the CMF surgery, because it looked to us like a like a good case to try out or prove our value. That is pretty common. And then we tested with this community we work together with with a hospital in New York and in Switzerland. So we arrived with this module. And I think, in the meantime, we got a lot of requests from surgeons that they would like to go one step further and not just, you know, to print out a specific part of the anatomy, but also to design guides. And I think one of the important surgeries where guides are really critical is the one that concerns the spine. And then our collaborators were really pushing us to, to develop those modules to tell them what kind of workflow they could use to to build their guides. And then after a couple of like successful surgery there we went together. That's, that's what we put our focus on. So probably next steps will be some cutting guides and and in some implants for examples calling plan. Yeah, let's see. So I would like to cover many procedures as many as possible. Of course, we have to prioritize, but Yeah, I do. I do modules that are coming soon. Yeah, so one of the common software that I encounter and my colleagues encounter is mimics from materialize, and they focus on a variety of organ systems, but also, you know, organ system based not like I can do everything, but they focus on heart, for example, soft tissue. Is that something that you guys are also going forward with in terms of vascular soft tissue modeling? So currently, we are really focusing on hard tissue like, like bone bone on CT and MRI. And for the, for the trauma coronial and orthopedic domains. We had some users that are we have some users that are using our software for for the longer, for example, for the heart, but I think it's not the majority of our users. Have another question here. I think it's similar eyes as a rhino medical makes automated rendering to ease the process. But does that affect accuracy? rendering or sorry? Basically, I think a segmentation is what he meant. How do you make sure that you're accurate? Let's say if I'm using your software, yeah, I got a motto. How do I know that my motto is accurate? Well, I think there's one part that comes from the algorithms that are behind and second, then the model that you're printing out is really like a one to one size. So you can compare it, you know, with your CT images, you can measure it. You can see also during the surgery if it's if it's pleats or not, but it should fit. I also have How do you measure if mimics is correct? I don't know. Um, I think this is definitely ongoing conversation as to how to validate your accuracy from segmentation to to print. Maybe there's some kind of, you know, those not guides. I'm blanking on the name of it, but calibration tools. Where are you are we really doing to try to compare, for example, our software, for example, with mimics, in terms of precision, because we mix is already like FDA approved? So we trust that the quality that it's producing is correct. And then what we are seeing in our studies is that we match up with this quality. Yeah, actually, with the orbital floor, I think we are even better because in terms of segmentation, for example, this is a hard problem, because the orbital floor is very thin, then you have to be able to capture it, why you, you have to do some extra maths that comes from, from the PhD studies we have. Yeah, I think these little finer details will be important. Once more and more people are using it, I think that morbido floor discussion is is pretty relevant, especially when he comes into contact with all kinds of bums. So that there's another question that's also relevant to our ongoing discussion about accuracy is do you guys, are you going through some kind of accreditation or some kind of governmental or like FDA clearance? Or c mark, clearance? That kind of thing? Yes. So our plan is now to conclude on the clinical trial, and then determined to reach clearance, we should go first. Because we have one in Europe, we have you have one in the US. And there are some other countries that could serve as the potential market for us like brisby, OBD, and visa. So we have to see where we go first. And then we will try to go grow from that. But it's certainly something that we are jumping in very, very soon. That sounds great. So I had a conversation last week, that was very interesting. Speaking of design for 3d printing, just because you want to make something doesn't mean it's going to print out the way you want it. And the mechanical properties also going to change depends on how you design it. What kind of features does your software offer so that whatever printer we use, either as an SLA or FDM, or maybe even higher end printers, what do you offer for different people with different printers and materials. So what we can guarantee is the closeness of the surface. So this is something that you will be sure that if you send your guide or your anatomy to the printer, it will be printable, this is a relevant question. And I think if you buy your like relatively good printer around like $5,000, there will be no risk. So I can suggest you some good SLA based printer, I don't want to advertise that accompanies. Also, we filament based printers, if it's possible to print multiple filaments, then we do support handling, it's much easier, because then you can remove the support very easily without you know, actually causing damage to your model. And this question somehow is very interesting to me. So recently, I purchased the $300 printer, just to see that, you know, really the limitations. And I was very much positively surprised, let's say in terms of usability and in terms of the printability of the structures, and what we are trying to give us an advice for the workflow is really something that to help people design guides that will be printable, even if, you know their means of investment, let's say is very limited. Well as design, you know, for example, if you designed this specific thing with with a guide, you know, if you design like shapes that are really around these and very, you know, designers then probably with a cheap SLA printer, you will have more difficulties to print out. But if you you know design like a rectangular shape here, which is flat and then you cut a little bit here like you You make it flat the cubes, for example, then even with the cheap printer, you can just perfectly print it out because it will be you know, on your on your bed sheet and then you know build plate and then it can grow the model and dimensions and such that even without the support you can you can print the tubes relatively. Well, last week, our conversation mentioned Oh $100 3d printer. Apparently, he was doing well with $100 printer. So I think the price is going to go down. I spent too much money. Yes. Well, I'm going to shift the angle of the conversation a little bit before more questions coming in. If we haven't any more. It's about your entrepreneurship. I'm kind of, I'm not very familiar with the European startup scene. But I know that you've been around for a couple years now, as a founder, what are some of the major challenges that you think are unique to European startup compared to you know, someone in the Silicon Valley, I think it's maybe a little bit mentality about taking a risk. And especially when you're in Switzerland, you, you start to learn that what is share is something that you must trust, and that's good. And what is risk is like, let's think about it twice, I think probably in the US, people are jumping on new ideas much quicker. And they really like to go for this kind of adventure. But I think what's the same is that you really have to be committed, and you have to be a little bit maniac about your product, that it will change the world, or at least bits will bring make it a better place. So how did you how did you meet her? Sorry, gone, you really have to find a thing, this inner purpose, that they every day, when you wake up, you feel that your product is great. And then you have to go independently, if there are difficulties if other people disagree. And I think it's the same in the US or in the EU in Europe or wherever you are, you need to be dedicated and motivated. Why but it sounds like if you're in Europe, you have to have twice the amount of passion to get things going. Because there just not a whole lot of cheerleaders around. Or that would that be accurate, it starts you know, I think over these couple of years that I'm around, I see that there's a positive change in terms of the startup culture is also investors are really getting the appetite, let's say to, to put money in startups and to, to try to shape the future. And I think now if you think of all these demographic trends, about the aging of the population, or the climate change, etc, you see that you have to invest now such that you will have the solutions when it's really trying to collapse the word on you. So I think it's, it's getting better, it's not there yet, like, you know, huge investments going on companies that are either collapsing or skyrocketing the next day. But I think it's it's a pretty, pretty lively ecosystem. So we also got a lot of support a lot of fun, there are institutions within Switzerland that are doing a cheerleading. So. So we felt a little bit like celebrities in this in this small ecosystem. But yeah, it's I think it's a little deeper. And also in terms of workforce, I think, in the US. If you work for a startup, you know, what you are engaging. In Europe, it's I think it's still coming. But there's a lot of young people there, I think are motivated to do something meaningful, something useful, and not just, you know, work in a feature number 20,000 of a product, but really well they have an impact. And I think it's it's pretty good. And I really like I when I did my PhD, I was teaching students, and they really, really love this mentality that now they are. Yeah, no, it's awesome. And also, you're a female co founder. So I just want to give you a kudos. I'm looking forward when you guys are starting to raise your, your your next round of funding, and happy to reconnect and have another conversation. And also, I realized that you have a beta download for is that a free version? I can just try it out, are you here we have a we have a better version online that you can you can download. And then yeah, we can have conversations, we are very open to collaborations. So if you have your medical data, and you don't really know how to process it, or you want to get involved, we don't without clinical trials, then you can send an email to our support address, and then my colleagues will jump on. Awesome. Savannah, thank you so much for spending time with us. And thanks for everybody joining us and asking these great questions. The recording is always available on igtv. But we will post processing into podcasts and etc. for distribution purposes. So thanks again. And thank you for having me. And I will keep you updated. Bye bye. Bye. Bye. That's it for this episode. Be sure to follow us on Facebook, Twitter and Instagram as videos and check out the links in the show notes. See you next time.