Disclaimer: This transcription was generated using ChatGPT. While efforts have been made to ensure accuracy, there may be errors or omissions. Please verify the information independently if precision is critical.

Cindy: So thank you again for coming to our talk. I’m delighted to welcome everybody today to this session on the intersection of artificial intelligence and women’s health. My name’s Cindy, as I have introduced. I’m the head of talent acquisition here for Rangel. We are really in a pivotal moment in healthcare and particularly within AI as it’s reshaping the landscape of women’s health. From personal diagnostics, innovative treatment options, the potential for AI to revolutionize the way we approach women’s health is immense. The integration of AI into healthcare has the power not only to improve outcomes, but address longstanding disparities with challenges faced by women accessing quality care. So today I’m honored to introduce our panel of experts and their perspectives and experiences in leveraging AI to advance women’s health and their experiences in AI and femtech as a whole as well. So without further ado, what I’m gonna do is have each of my panelists do a bit of an introduction on who they are, and then we will go into the questions. So I will start with Garima.

Garima: Thank you, Cindy. Good evening, everyone. My name is Garima Gale. I’m an academic scientist by training, biophysicist to be precise. I’m passionate about bridging the gap between science and business. As a senior program manager at Ontario Bioscience Innovation Organization, or as we like to call it, BIO, I lead new programs and initiatives to support early-stage and scaling health and life science companies. Currently, I’m leading our Women in Health Initiative. This is a program designed to support women entrepreneurs in health and life sciences. Really looking forward to our conversation today and with you all. Thank you.

Shazia: Hi, everyone. My name is Shazia. I am the founder of Ocare AI, which is a dental AI company here in Toronto. And I have over 12 years of experience building machine learning algorithms for the healthcare space. So I worked in oncology, pathology, radiology, and now dentistry. So I kind of specialize in building models in those spaces. And I am super interested to see how we’re going to incorporate women into all the tech that we’ve developed today. I want to see way more products than we see right now in this space. Loads more. So super excited about this talk. Thank you.

Emma: Hi, everyone. My name’s Emma Tolstorf. I’m the co-founder and CEO at AI-MD. Like these lovely ladies, I have a PhD in neuroscience, but mine isn’t neuroscience. It’s in something other, scientifically challenging. But yes, my company’s called AI-MD. It’s an AI-driven symptom-checking tool that allows people to scan themselves for disease from their cell phone. So essentially what we do is we harness the hardware on the phone, say, you know, like the sensors, the microphone, the cameras, in order to collect information, in order to diagnose diseases on humans more readily. And so we’ve had this company going for a few years now, and we’ve begun our commercialization, and it’s been a very exciting process. And so machine learning and AI is near and dear to my heart as a, I guess, a software engineer. And I think that it has a lot of places it can go, and it all starts with these types of conversations. So I’m super, super excited to be here today. So thank you, Cindy.

Cindy: Absolutely. And I’m super intimidated. I don’t have a master’s or a PhD, but I have street smarts, so. This question, I’m gonna start off with a question for everyone, so feel free to chime in. But how do you perceive the current landscape of AI-powered healthcare, specifically in addressing women’s health needs? What are some of the notable advancements or innovations that you’ve recently observed? So I’ll let, is anybody interested? Garima.

Garima: I could start. Yeah, that’s a great question. And I think I would like to start this conversation by maybe framing what’s women’s health, because traditionally women’s health has been very much focused on reproductive health. But over the past few years, we have seen that change to incorporate a much broader perspective. So typically into three categories. First is, of course, the reproductive health conditions that relate to fertility, maternal care, contraception, and so on. Second are conditions that are specific to women’s biology. So cases like the menstrual cycle, menopause, and so on. And third, which has been in the past neglected, but now slowly coming up is conditions that are either disproportionately affecting women, like diseases like autoimmune diseases, or differently. For example, cardiovascular diseases. The symptoms that you see as a heart attack in a man versus women, they are different. And in many cases, what is a symptom for a heart attack for a woman is often diagnosed as a heart one. So keeping that sort of framework of women’s health in mind, in terms of recent AI-enabled technologies, I can speak from the companies that we have engaged in. So we see a lot of AI in terms of fertility and hormonal app tracking. So for example, Future Fertility, Juniper Genomics, are a few of the Canadian companies working in AI fertility space. There is also a company based out of Quebec, Eli Health, which is pretty cool. What they have is a device, wearable device. What it does is it provides you your sort of hormonal cycle every day, and it’s simply from a saliva test that you can do at home. We also see a lot of companies around pelvic healthcare, like Cosm and Hi-V. Those are, again, companies here out of Canada. So just summarizing that it’s great to see how AI has been enabling this technology, specifically to see that these healthcare are being more accessible to women in a more consumer-centric fashion, and in a way that it’s also addressing areas that were traditionally underserved, like pelvic healthcare, menopause, and so on.

Shazia: That was a good answer. That was a really good answer. I don’t know how I’m gonna follow up. Okay. Since you did a great job on the healthcare, on the women in healthcare, sorry, sorry, I’m gonna focus more on the AI side. So I think what I really love that’s changed in the AI field is before we used to dump all this data into one machine, and we had no idea what went out, we had no idea what came out, but we assumed we could apply it to anyone out there in the world, men, women, any sex, any age, any gender. But what I love right now is that people are actually curating data sets for women. They’re actually building products for women. Even if they have a product that’s for men and women, they’re actually putting a data set aside for women because we know that women’s behaviors are very different from men, even in mental health. Women’s mental health works very different from the way men’s mental health works. And I think it’s really important to recognize that and actually plug it into our system, plug it into our tech. So I’m actually looking forward to seeing more of these personalized systems in the tech space. In terms of the companies I’ve seen, a lot of them you already mentioned, so I’m not gonna re-mention them, but one that I came across recently, which I love because they’re thinking about not just the individual processes we go through, but how we get there. So they were introducing egg freezing into our benefit system. How do we get our companies to recognize that one day we might wanna freeze our eggs? That’s just a leg up. It’s getting us to a place where we can even consider stuff like that before we get to that state, which I absolutely love. I wanna see more of that, actually. I think there’s a lack of those kinds of technologies that lead up to these individual cases.

Emma: Yeah, and I think, too, there’s a large proportion now of companies that are getting into this diagnosing things before they happen, right? Looking at things like trying to predict before the problem happens, not answering the question of what do we do once this problem emerges, right? And I think there’s a lot of what we can do is highly predictive in nature and is very exciting. And I’m really excited, personally, about the emergence of IoT and devices becoming integrated with these types of systems so that we can collect data, even if we don’t know that we’re collecting data. It’s being collected in the background. It’s, you know what, my mom, I track my mom when she turns the coffee pot on in the morning. I do, personally. It helps me to gauge whether she is not feeling well. And I also have other Google sensors in their home to help me know what’s happening. I’m not weird. I’m not completely weird. But essentially, if I notice a delay in the time of her coffee pot turning on, I start to get an inkling that perhaps my mother isn’t feeling well. And so there’s other aspects like that that we can start to train, like computational models, to start to understand the tendencies of human beings and be able to figure out what is not normal for any given person. And it’s not a one-size-fits-all model, right? It’s not a, we can’t look at this group of patients the same way anymore. And so it’s allowing us to actually deliver personalized care properly, the way it was intended, the way human biology was actually created, right? So I think that’s the most important and the most exciting aspect for me.

Cindy: Wonderful. Thank you. Thank you. Awesome. Shazia, this question is for you. So personalized medicine is often touted as a major benefit in AI and healthcare. Considering the unique physiological and hormonal aspects involved, how do you envision AI facilitating personalized approaches to women’s health?

Shazia: Yeah, so I love personalized healthcare. Like I think we could personalize to the individual as Emma believes with

her product as well. I think there’s, we each have unique characteristics that should be captured in any kind of dataset or product. And I love that I can go to Sephora today and not only have makeup that matches my skin color, but they also look at the texture and quality of my skin as well and give me products for my skin. So I just think that healthcare is going this way, not just healthcare, but all these other tech companies as well. In terms of how we get there, that’s a really good question. We need data to get there. But I also understand that some of you might be starting a startup. You definitely don’t have the kind of money to go out and actually curate these datasets. So even just starting to ask your customers the questions about how their processes are different from their husbands or their partners or whoever they’re living with, actually asking the questions to try and find out what unique differences there are between your customers. It’s a great way to start because then when you go to collect data, you can actually start to target where you collect that data as well. But yeah, personalized medicine is the way to go. Personalized healthcare is the way to go. Because when we recognize what’s good for us as an individual, when we recognize what products are really good for us, what tech is really good for us, what medicine is good for us, we get empowered as women. Next time we go for a product, next time we go for tech, we know exactly what we want. We don’t need to ask anyone. I want us to be extremely empowered to a point we don’t need to turn to anyone for any information or advice. That’s where I want us to be.

Cindy: That’s wonderful. Thank you. Thank you. This question is for Emma. So access to AI-driven healthcare solutions can be varied based on socioeconomic factors. How do we ensure equitable access to AI-powered healthcare technologies for women across different demographics, regions?

Emma: Amazing. I’m super excited about this question. I always love this question. I always pick it out in the list. So for me, I love this Sephora aspect that you’ve found. I think that is the absolute future. And I think the first way that we can make things more equitable is by actually including enterprises in our business models and getting away from subscription models. Because when we start charging a bunch of subscription models, then that makes things not equitable. And that makes it so that we are a subscription jungle and no one can really purchase any of these tools that we’re developing. So what we really should be doing is we should be partnering up with these enterprises that have this huge marketing and sales budget and want to move their products. But we shouldn’t be saying, so say you had a system that could tell you what skin type you had and what ingredients you should have in your skincare product. Instead, you should, as a business owner, go to someone like Sephora, like the brands at Sephora, and say, hey, I want to use my AI and then I want to push your products for a commission. And then that way, I can offer this to consumers for free. So consumers get to benefit from the technology, but they aren’t actually the one paying for it, which I think is going to be really important in keeping things equitable. Because there can be win-win scenarios for large enterprises with the budgets for this stuff and for consumers to use as well for free. And, well, for free. It’s, right, the enterprise pays. And then the other way is to harness hardware that people already have. You’d be surprised what you can come up with with things that you probably already have in your house or in your purse. At AI-MD, our system operates completely from a modern smartphone. The technology was actually inspired by someone who I knew in Mexico who went to Mexico City to start this company called Phone Drone. Phone Drone, it is a drone company, but instead of, oh, I should actually preface this. Drones about, what, seven years ago were really expensive because they require cameras and sensors and accelerators and a GPS and all of this hardware inside the drone. And not everyone could have a drone for that reason, right? It was so expensive. But my friend went and he created this company called Phone Drone, and it allowed people to set their phone inside the drone so it harnessed the hardware from the phone and powered the drone. And it was really awesome. They broke a ton of phones, but it was really awesome. And I like the concept. You know, the concept is that your phone, right, contains a lot of really great hardware that can be medical diagnostic in nature, right? And that’s what inspired the AI-MD invention and being able to carry through all of these diagnostic technologies. Because, you know, let’s say, let’s look at the thermosister in the iPhone, right? I just want to spark some creative juices here. We have something called a thermosister in an iPhone, right? And essentially what it does, it tells you if your phone is too hot when you’re at the beach or if it’s too cold when you’re ice fishing. And it says you better get this phone in a better environment or there’s gonna be a problem. And so that technology can actually now be used to take a body temperature on a human being, right? So that now this person who doesn’t have money for purchasing, you know, a thermometer can actually take their body temperature and can actually be in control of their health. And that’s gonna help with all kinds of things like fertility, right? Which is going to be a huge problem. And so the beauty too of the cell phone and putting things on the phone is that in a rural community in Africa where they have a barely a house and like maybe not even a toilet, somehow they still have an iPhone. So that’s really important for making sure not only equitable technologies can exist but also global technologies that are also equitable can exist.

Cindy: Wonderful. Thank you. Just a thought just came across, you know how we said all the informations are in our phone or it’s like whatever wearable device we wear. But just thinking to build a few questions, is there any startup or technology being developed? I mean, we have our hormone tracking, temperature and so on. Let’s say, you know, I’m on a diet or whatever diet it is and I say, okay, scan the food or, you know, give the details on the food and says, well, maybe not today because your cortisol level is high or low and what do you think?

Emma: Yeah, well, absolutely. With AI-MD, that’s a great question. So essentially, since cortisol is a tier one hormone, it can be related to something called heart rate variability. Who knows what heart rate variability here? Yeah, it’s a really cool thing. You should look into it. So heart rate variability can actually be utilized to predict a variety of biometrics of the body. And so essentially what we can do is we can take a 30-second video of your face, measure the blood flow in your face, extract that data and utilize like the heart rate variability metric in order to predict how much cortisol is in your blood. Today, I was a three on a five-point scale. If anyone wants to give the app a try, I’d be happy to share the details on that. I need to get one of these posters. But yeah, that’s essentially how that would go down.

Shazia: Well, I would try it, you know. I tried to create an app that could take a photo of food. Food is very complicated. Once you cook it, it transforms into completely something else. I couldn’t find a way around it. So if anyone else builds it, I want it.

Cindy: This is a great segue to your question, Garima. So looking to the future, what do you envision are the most exciting possibilities for AI in advancing women’s health and what we’re calling femtech? What steps can stakeholders in healthcare and tech to accelerate the progress in this area?

Garima: Okay, so, okay. Since you mentioned femtech, so it really got me started. And it’s like, where did the term first come from? So, okay. To the audience, does anybody know when the term was coined or by who? It wasn’t that far along. Actually, it was in 2016 by a co-founder of an app called Clue. I don’t know if you guys have used that. It’s a menstrual tracking app by Ida Tin. I think she’s a Danish entrepreneur. So it’s in 2016. It’s what, 2024, eight years. And the way femtech women’s health has revolutionized, I think it’s great. So back to your other question, where do I see? And this really goes back to both of our panelists’ point on precision, personalized medicine. So it’s been touched upon. The thing is, the way a man’s body or a woman’s body or personal body composition reacts to drugs or devices or treatments, it’s very different. And historically, what we know is the medicine or the treatments that we develop are all based on male physiology. So if, let’s say, not until 19-something, women were not included in clinical trials, right? And for example, pregnant women, they don’t necessarily go in clinical trials. And also, there isn’t a proper animal model, for example, to study what happens during pregnancy. So all coming back to the point of precision medicine. So where I really see this AI in women’s technology, but this is, again, of course, the caveat that there is similar progression happening in other sectors of science, is a, how do I call it, like a digital twin, right? So digital twin to study pregnancy, digital twin to study endometriosis, right? So I think that’s where we could go if AI and rest of the sciences sort of work out. So that’s where I see it going, in

terms of what stakeholders and rest of the community could do. I think this, again, is we really don’t have much data. So even with the drugs that go through clinical trials, you have, for example, let’s say 50% men, 50% women in a clinical trial. What you don’t have is what are the effects of the drug on those populations? Because the effect of drugs are different based on whether you’re male or female on a broader level. So I think in terms of what we could do, the innovators, scientists in the community could develop innovative tools to really study female, right, you know, female biology, female physiology. And in terms of technology, I think it’s conversations like this that need to happen so that everyone’s aware. And if we keep this momentum going, I think the digital twin for, you know, female-specific diseases or conditions, it’s not far, that far. That’s my parting thoughts.

Cindy: That’s wonderful, wonderful. Shazia, this question’s for you. Algorithms are only as good as the data that they’re trained on. So how do we address the historical lack of diversity and inclusivity in medical research data to ensure that AI-driven diagnosis is effective for all women?

Shazia: So I reluctantly pick this question. Because medical research, like medical research takes so much time to do. Oh my God, you have institutions, you have politicians, you have hospitals involved. It takes so long to do. And, you know, for a very long time in those universities and hospitals, the people in charge were typically old white men. So they didn’t really care much for women’s health. So as a result, our data is very, very biased towards men. And it’s still like that today because it’s a very old system and it’s very hard to change one of those organizations in the way they think. So I do challenge you actually to change the way that they think, partner with them, get some new projects going. You know, I came from academia. I finished two postdocs before I left academia. And one of the things as a young researcher, if you’re trying to get into a professorship, you need to bring in funding. And if you can partner with a company and bring in funding, people suddenly love you in that institution. Imagine if that funding actually served the women that you wanted to, you know, change their lives for the good, for the better. I mean, that would be incredible. So definitely, if you see a young female researcher that’s trying to pave their way, help them out, give them a hand, you know, have a chat with them. They are actually looking for help to get into these really high positions in those institutions. In terms of the bias, oh, the other thing about medical research, I wanted to tie into something you said about clinical trials. A lot of women do not get into clinical trials. There’s a lot of exclusion criteria. Like if you’re pregnant, you’re not allowed to get into this clinical trial. That’s one of the big ones. And usually it’s for a good reason, but I feel like sometimes it’s just put in there just in case. So you just, you’re never gonna get data from women who are pregnant. So that’s something to definitely take into consideration if you’re taking that data and using it to train your models. And the second thing I wanted to mention about the bias, the AI industry was built by men and the data was for men and therefore the models and the products are for men. So if we really want to change it, we need to start collecting our own data. We really do need to start doing that. We need to tip the balance where it’s now like 90% data from men. We need to actually tip that balance the other way around so that we can kind of add on top of that data. If you are collecting data, I would advise kind of seeing what’s out there already and trying to fit into it because you’ll find more adoption if you do it in a format that’s already there. Whereas if you’re trying to bring in something brand new, you might find the adoption part really hard actually. So yeah, try and gel with what’s already there rather than try and force your way in because that might be a little bit difficult. But yeah, and also know who you’re trying to collect data from. If you intend for your technology to work in Africa as well as in Toronto, as well as in Europe, be very aware that some of them don’t have access to expensive equipment and maybe all they have is a phone in their pocket. And maybe adjust your technology so that you can get those colored women across various ethnic backgrounds into your datasets because they won’t have access to all the hardware in the world. And that’s my two cents on that topic.

Cindy: Wonderful. We have one more question and then I’m gonna hand it over to everyone if you have questions in the audience. So this one is for Emma. So cultural and societal factors are often influences in women’s health-seeking behaviors and attitudes towards tech. How can AI-powered healthcare solutions be culturally sensitive and inclusive to effectively engage diverse communities? So kind of along the same lines.

Emma: Absolutely. Yeah, no, I think AI, it eliminates the embarrassment and the humiliation potential of being exposed to a human being, you know? When you have a problem, specifically in healthcare, right? There’s certain things you wanna say to another human being and certain things that are a little bit more, or a lot more painful to say to another human being, right? And I think when we have AI, it creates this sort of veil of secrecy and it allows us to keep our, the weird things that happen to us, to ourselves, but it helps us to make understanding of them, right? It helps us to make sense of what’s going on with our body and what the next steps are, so to speak, right? So I think largely that has many societal, or yeah, societal and cultural implications, because I think these, I guess, I think different diseases perpetuate in different cultures because there’s different cultures that are more progressive and are, you know, they’re gonna tell you the whole story, the whole truth, and nothing but the truth, and then there’s other cultures who, there’s a lot more shame and stigma associated with, right, presenting a certain symptom set, so to speak, right? I don’t know, I was thinking about this story, but I’m gonna tell it. So I had a friend, it was a couple of years ago, and she called me up one day and she said, and to preface this, a bit of context, I’m the person that people call when they have a problem. And I’m the solutions person. And, you know, she said, Emma, I have a pregnancy and I need to be not pregnant. I said, oh, okay, all right, well, you know, well, you’re gonna go, she’s like, it’s been, you know, 14 weeks. I said, okay, well, that’s just fine. You’re just gonna go to the doctor and you’re just gonna tell her that you want to be un-pregnant. And so she did, and she got this woman who was not culturally matched to her. And so in the clinical space, I started to look into this after this had happened, but the research says that when you are working with a provider that is not culturally matched to you or racially matched to you, or not sharing like the same values as you, it is much harder to receive proper care and overall patients rate their care to be worse than it is. And so essentially my friend went to this culturally different physician and the physician made her feel very, very guilty for her position, was not willing to help her, made her go explain her whole story to someone completely new. And it was like a completely like terrible situation, right? She was in a pickle to begin with, right? Without this having to go to multiple physicians who maybe didn’t agree of her choice in course of action. So I think certain things can be really uncomfortable and like hard to get through when you have to bring it up to other humans. It’s hard, right? It’s hard to share things that are shameful or embarrassing. And I think AI can really bring us a long way. They can help us to be more informed. AI could have helped her in the early stages and she could have said to a large language model, hey, I’m 14 weeks. Hey, I have this problem. Hey, this is the symptoms I’m experiencing. Am I hooped? Can this actually happen? Am I gonna have to like give birth to a baby that? And it could have given her some insight into what was going on prior to her going to the doctor. So to avoid like a full-on nuclear adult meltdown, right? And so it’s these things that can help. It’s honestly, sometimes I feel like I don’t know if I’m in like the diagnostic space or the anxiety reduction space, because that’s truly is like the outcome of these tools. And so it is very interesting and very, it’s very, very exciting to see the, I guess the potential for these types of technologies to be helpful in reducing and mitigating concerns related to like cultural and societal aspects.

Cindy: Wonderful. Thank you. I’m gonna pass it over to the audience. Does anyone have any questions, comments, or anything in the audience that we’d like to tackle?

Audience Member 1: I do have a question, but I’m not sure who can address it.

Cindy: Okay. Sure.

Audience Member 1: Regarding the data for AI and machine learning for training the algorithms and stuff, how safe is the data? How do you see collecting the medical records or symptoms data? Like for the reference of what’s going on with ChatGPT, OpenAI, and Google Gemini, apparently they train the algorithm, the text language algorithms on YouTube and other things that they were not trained on. They were not supposed to, which is borderline illegal

, but that’s just text and videos. Now we’re talking about medical records. We’re talking about patients. We’re talking about symptoms and like treating people. So what do you think about the data in that space and how secure is that and how progress is gonna continue?

Emma: There’s a lot of data that you wouldn’t wanna steal. EMR is like traditionally have very bad data, right? Because it’s recorded by physicians who are not researchers, who are trying to put down as much and as little information as possible to keep a running record of what’s going on with that patient, basically trying to ride the line between getting sued and being able to advise them when they come next, so to speak, right? It’s a hurried and rushed process. And the detail in the notes aren’t stellar to begin with. So I think it’s important too to think about whether I guess we would be wanting to take the data in a lot of cases, but maybe, can you speak to this one a little better?

Shazia: Yeah, I mean, how safe is your data? It really depends on who you’re giving your data to and how they have explained their system to you. Like I’m a dental AI company. We don’t take photos of faces or anything. We only take photos of teeth. So it’s fairly anonymized. We don’t know who you are. So people are very safe. They feel very comfortable giving us their data. But if you’re scraping your age, your sex, your gender, your family history and stuff, you probably wanna ask, so how are you storing the data? Are you gonna store my name without data? Do you need my address? Ask those questions, because you’ll feel much more comfortable about where your data is going. I think with the EMR stuff, I do agree with everything you said. However, we need to validate our model somehow. And unfortunately, EMR records are our ground truth right now. They’re the best we have, these physicians writing crap down. So in order to convince you that our models are working, we actually have to say, these models perform very well with these EMR records in comparison. And da, da, da, da. And this is how we know it’s really good. And this is what it does. So yeah, the EMR records, major problems. But I think for a lot of systems, it’s like our only ground truth. Like I worked on mental health system. And the only ground truth they had were questionnaires that were filled in by the patient themselves. So literally ones, twos, threes, fours, fives, note it down on a piece of paper. That’s a ground truth. They don’t know how else you’re feeling if you don’t fill in this questionnaire. So yeah, no. Medical information is, it’s safe depending on who’s actually looking after your data. And there’s no guarantee. Someone could come along and just steal your data.

Emma: So our ground truth is in a pretty sad state of affairs.

Shazia: Yeah. But yeah, right? It’s encrypted also, right? In transit and at rest. And essentially there are ways, when you’re transferring data to the cloud and back down to the device to secure that process, there is other fields. And so we follow similar protocols in other fields as well. So that can help. And also the de-identification and the randomization.

Emma: Oh yeah, 100% anonymized. Even if someone takes it, they shouldn’t know who you are.

Shazia: Exactly. 100%.

Audience Member 2: Yeah. So I believe there’s HL7 and other protocols, right? One of those protocols.

Emma: Yeah, like the FHIR protocol.

Audience Member 2: Yeah.

Emma: And so you can create your API around the FHIR protocol as well. So many, sometimes a healthcare company will have like an API, standard API, and then they’ll have like the FHIR, like HL7 API as well.

Cindy: Yeah. And again, give two perspectives. One from the side of companies who have worked on this space. For many of the companies that are using data and want to either incorporate with healthcare systems here in Canada, and anytime you give out your data, they do have to have HIPAA compliant, right? And one of the reasons why, especially Canadian hospitals and networks are so averse to adopting these technologies is, first A, it has to be integrated within their system, and second is, there’s also the security risk and the data risk, because the minute another technology comes integrated into their system, they have another layer of data risk. On a personal note, and I might be the one in this room, if having access to my personal medical data, like imaging and so on, gives me access to faster diagnosis, faster appointment date, I mean, would I care? I mean, right? I mean, right now, how long do you have to wait for like basic, I don’t know, screening? Three months, four months? I don’t know, just a thought out there, and I may be naive.

Audience Member 2: Yeah.

Audience Member 3: Okay, great. My clients had a data leak in 2019, or when was it? Now they’re reimbursing for Deloitte.

Cindy: Yeah.

Audience Member 3: So that was a huge one, because it’s like blood and…

Audience Member 3: So they’re very careful not to adopt the technologies, like, you know, especially like healthcare hospitals and it’s in general, they’re very, discriminate a lot.

Cindy: One more. We have another question here.

Audience Member 4: Hi, my name is Asha, and my background was in research, and thanks a lot for coming today for this panel. I have this interesting question. So you were talking about how slow medical research is, and what we are trying to do here is that we wanna try to utilize AI and digital health to help women better. Now, coming back to medical research, if you want, I don’t know how true is this, but apparently, like even our female genital anatomy and physiology is not clearly illustrated and described in medical books, and God knows what we are teaching the physicians. Like there are 10 pages on penises and one paragraph on the clitoris and cross section of the, you would understand, like of nerve ending, like only right now we are having studies of how, what nerve endings are there in the clitoris, what clitoris even looks right, and when people do surgery, what they’re doing it to. Just going back to when we are trying to link this deficit in medical research and what technologies, we are always trying to optimize healthcare, right? Give you better, give better health information, but if the deal, if the lacking is in the base, like in the core information and knowledge that we are giving, how, like, do you find that the solutions that you’re coming up with will fall short of really executing to their best optimal promises because we are lacking in that information and.

Emma: How to build a house on top of a sinkhole.

Audience Member 4: Yes, yes. I have a lot more questions, but this is the main question. Maybe I’ll find you out later in the, and ask you this question, other questions, but this is my main question about how, like do you feel like sometimes you are just again, like, you know, forming something based on, like, you know, data that is lacking, information that is lacking, and how do we address that now?

Emma: Well, we address that by not utilizing data that’s so subjective, right? So physician notes, for example, I have a real beef with physician notes, right? Because it’s highly subjective. You get, you know, Dr. Bob, you get Dr. Jim, you get Dr. Rob, and they’re all gonna say something a bit different about a patient. It’s all gonna come out a bit different in the research, and the core problem is that you’re getting another human’s observation, right? And it’s not actually what exists, right? So we actually need to be collecting biomarkers, right, with devices that can collect biomarkers, like a cough sound, for example. For COVID-19, we could collect, and we can say, okay, based on this cough sound, we can say, okay, this person has COVID or does not have COVID. It does change your cough based on, you know, the presence or absence of COVID-19 in many scenarios. We can do the same thing with pneumonia. We can do the same thing by taking a photo of like a skin condition, right, and saying, okay, that is atopic dermatitis. This is melanoma, right? And we can actually differentiate diseases based on objective biomarkers. So we get rid of this subjective funny stuff, and we replace it with actual objective markers that will be good today, and it’s gonna be good tomorrow, and it’s gonna be good 100 years from now, yeah.

Shazia: Regarding the knowledge, we just need way more females in those education systems. Like if you shoved a million females into U of T right now, how many of them are gonna work on that kind of research? Probably a lot of them, actually. But they’re not getting funded. They’re not getting any money from the government. They’re not getting any money from the institutions. And we’re just like, why is this not happening? Well, there is a solution. Just throw money and people at the solution, and you’ll probably get there. But we need people to care, and the people at the top need to care. So yeah, that’s how we’re going to get better knowledge on those areas. It’s not hard to get the knowledge. It’s just someone has to sit and get the knowledge.

Audience Member 4: Can I just add to that a little bit? But exactly, there’s knowledge is there, but

people are being very stubborn about adapting to the change in information, or even wanting to understand more about even orgasms, for example, that is going to be found, but it is important. It is important. I watched a three-part episode on Netflix about orgasm. It’s really fascinating. But it’s a very, no one wants to care, or we have so many, I’m just thinking, people who don’t think this is important.

Shazia: Yeah, it’s really weird, actually. It comes back to your point about being embarrassed or something. It’s the same in academia. Women are embarrassed to go into research about our reproductive systems. It’s really bizarre.

Emma: Well, it’s the same under the school, right? I’ve actually asked about this specific thing. It’s like, why isn’t there more orgasm research? You know, like, what is going on here? Right, it’s something that we all do, hopefully. And, you know, and if you’re not, then that just goes to show there’s not enough research here. But, you know, it bestows shame onto the university when you are researching those things. Universities are extremely conservative, right? So when you try to innovate, when you try to, you know, be like, okay, this exists, right, it may not be, or it may be too sexy for you, right? But yeah, it is shameful for universities, I think, too.

Shazia: It is. I used to work on breast cancer research, and I used to put literal breasts on my slides. I used to hear laughs in the audience from full-grown men. I’m like, oh my God, get over yourself. Holy shit.

Audience Member 5: All right, we have time for one more question, and then we are going to leave room for everybody to do networking and to have these continuing conversations. So we have one here.

Audience Member 6: Thank you so much. So my name is Anna, and I came from Russia, and I lived in four different countries. I lived in four different countries, like in my 30s. So my question is about collecting data and interpreting it from, like, different medical providers worldwide, literally, like from the CIS country to European countries to North American countries. And for example, like when I go to the OB-GYN, it’s literally like a puzzle game with Google Translation and sometimes even Google Pictures to literally show what exactly I have and what data I have. So literally, and I think that it’s a question from everybody here. So how fast can we create a tool that can collect, interpret, and do something with it from, like, to Canadian providers, for example?

Garima: I don’t know, I’ll just share the same problem I had. I moved with my daughter from Denmark to here, and all her electronic immunization records were in Danish. So, you know, there you go, I had to translate it. And where do I translate it, right? Google Translate, I don’t know if it’s correct, but, you know, try and find. To be honest, I really don’t know, because that would require, what did you say? I mean, you’re talking patients and from health and healthcare information. You’re talking all different, even within the US or Canada, it’s just one country, different provinces, different states, you have so much trouble streamlining it. Now you’re talking internationally. I’m not saying it cannot be done, I just don’t know how yet.

Emma: I would rely on the data, actually. You know, the imaging data, the blood tests, the lab tests? So the tricky part here is the language, right? That’s the barrier here, the language. So if you just remove the language and started from just the core data, each country had their own way of taking that data and transforming it into something that they could understand, I think you could get there, actually. It’s just how much data you’re willing, how much time you’re willing to spend to ingest all of that data and all of those lab tests. I mean, maybe you could do it, Emma, with your app.

Emma: Well, no, absolutely. Well, there are like mechanisms in place like to allow someone to do something like this, right? Like you can fully transcribe a document and you can have a photo of a document and you can extract the words on it, right? And then you can translate those words. And translation now is better than it’s ever been, ever. And so essentially, you know, utilizing large language models that are existing, translation schemes that are existing, it is becoming more and more possible by the day to develop something like that, right? And to have something like that and be able to offer and start understanding, you know, regardless of the language that you’re speaking.

Shazia: Yeah, I said build it, honestly. I could see so many applications. We just gave, my fiance’s Russian. We had to get it translated to get married later this year. You could use it for that as well. There’s so many good tools for that. I think you should build it.

Emma: It’s on the tip of my tongue. My AWS guy has a thing with AWS that actually does that. It pulls the information off. I’ll figure out what it is. I’ll call him. I’ll call him. We’ll be in touch.

Cindy: Okay, wonderful. Well, we’re gonna wrap it up, but I always like to leave with a final thought from our panelists. Do you have, would you like to impart some wisdom, some final thoughts, or anything that you’d like to add before we wrap up and start to network and mingle?

Garima: No pearls of wisdom, but thank you all for listening to us, and thank you Danielle and Cindy for organizing this. Looking forward to talking with everyone here.

Emma: I’ll impart a, I will impart a piece of wisdom. So back when I got going, I got going in my entrepreneurial journey, and I like to sort of highlight this for women because I think sometimes we’re like, ah, I don’t have enough money. Ah, I don’t have enough time. Ah, you know, the conditions aren’t so favorable, and it makes us not start, right? And the important part is really starting, you know? Because once you start, then you can pick up some momentum and get the show on the road. But for me, I remember when I started my first company, it was, I got out of my undergraduate degree. I got this really cool job in the clinical stroke neurology lab, which seemed to be fantastic because I had attended university against the advice of my parents because they were entrepreneurs. They said, Emma, don’t go to university. I said, no, mom and dad. I want to join this cult. And so I did. And so I finished my four-year undergraduate degree. I got my clinical research job, and the function of my job was to convince people to donate their cerebral spinal fluid in this stroke neurology lab. And I was the closer. I was closing the CSF like crazy. It was awesome. It was the best job. But then a couple of weeks in, I learned that the lab didn’t get their funding for my position. So now I was a girl with a four-year degree, which her parents didn’t want her to have, and I had no job. And so I was like, this isn’t good. This is very bad. And so I went home and I started to look around. My roommate was a little weird at the time. She ended up moving out. And so now I had this two-bedroom condo, no job, couldn’t tell my parents. It was a fiasco. So then what I did was I actually started up this Airbnb business called HostBot. And so essentially it was an Airbnb renting system, but I rented out her room, and I actually rented out my room too. And I slept for months on this futon couch that was like, it was literally the worst. It was like Ikea outdoor futon couch material, right? It was literally the worst. So I did that for a couple months, and the condo was 1,200 bucks, and I was renting each room out for 100 bucks a night between the two of them. And by the time the month was over, I had three grand. And I was like, huh, I should multiply this. So I did. I went out and rented a bunch more condos. And then I began to Airbnb those too, and this was back in 2019, before this Airbnb arbitrage. I invented, I assisted in the invention of Airbnb arbitrage. And essentially that’s sort of how I got going, and that’s how I got the money to finance AI-MD, right? So that was bootstrapped all the way up to like a $400,000 a year generating company, right? And so that is how I got started, on a futon. So the message I wanna tell you as the final piece of wisdom is that it’s never a good time, right? Be engineering and try to think about things in different ways. And you need to make the money in a different place because getting funding from external sources is less educational and more will come from it if you can get going a bit with resources that you can bring to yourself. It will reduce your imposter syndrome and it will increase your will to really find success and it’ll help you feel it.

Shazia: I don’t have an awesome story like that. I was just sick of working for guys. So I was like, I’m gonna build it myself. No, I think my last advice is if we wanna change this fintech fields and have more people talking about it, we need more of us in those CEO, CTO positions, right? We need to get more of us into those positions so that we can make the decisions about what kind of data we’re gathering and what

we’re building and what kind of product’s gonna end up on our app market space. So yeah, I would kind of what Emma said, which is just start, start doing it. We need you there. We really do need you there to be able to change the space so that more of us are in that tech space. So please, please do start.

Cindy: It’s a great message. Thank you. Thank you. Thank you. Thank you.

Cindy: Well, wonderful. I wanna thank the three of you. I appreciate your time and your insights. I think this was a really great conversation. So thank you to all three of you. Thank you to Women in Tech as well as the Firehood. And now we get an opportunity to mix and mingle and have some refreshments and continue the conversation. So thank you again.

All: Thank you.