Healthcare & Reflections

Key Takeaways

[Music] according to a recent study by the World Health Organization about half the world's population still lacks access to basic health care services at over 3.7 billion people for some it's because they live in a rural area out of reach of health care facilities and for others it's simply because they can't afford it but the internet is fast connecting every human on earth and since we've now seen how we can use it to offer people in the most remote regions of the world social networking services we have to ask ourselves can we do the same for healthcare services can we use technology to save lives at the same scale as say Facebook or Google to find out I gathered a video production team together and we took a trip to Tanzania to meet the brave entrepreneurs that are actually creating the software based healthcare services the world in Siraj we are here in Tanzania we are here to find some interesting people that are using AI technology to make some really incredible solutions for people everything from cancer detection to helping young girls with educating them on sexual health to all sorts of technologies that are used to prevent poachers from being successful like drones in the air that can protect humans the first person on the list to meet was Ali Salim an entrepreneur in the Kilimanjaro region working on a cancer detection app called dr. Elsa the goal of dr. Elsa is to utilize machine learning technology to empower doctors and institutions to provide smarter better health care for patients in East Africa I sat down with him at his office to have a conversation about his history and his work how's it going away good yeah yes it is so tell me a bit about yourself like did you grow up in Tanzania oh yeah absolutely so I born and raised here I went to university employed then I moved to Texas then went to California then move back yeah okay so that's a lot of so a lot of in the u.s. yes yes yeah so I said my background is in software mathematics I work for a small start-up email Los Angeles and then came back home because this is where I'm needed this is where you're needed yeah so you felt a call to come back to Tanzania rather than because you I'm sure you could have worked in the u.s. if you wanted to yep well yeah I just love it here so much and I miss my mom oh okay that's that's a good reason yeah sure for sure are you what's uh like what what's what's a day in the life of you like you what's what does a day look like for you typical day oh man recently they change around a lot so I wake up very early at like 4:00 in the morning I start working until like 8:00 a.m. just working on dr. Elsa and Mary so Mary's in you a project we're trying to push across East Africa but yeah so most of my time is spent between hospitals between meetings and talking to regulators and policymakers Wow okay so a lot of people don't think of that when I think of hey I their development but that's such a crucial aspect especially for your use case yeah that's the hardest bar - that's our technology is super simple compared to talking to people in the government and just pushing for new policies to be implemented yet can you talk a bit about the tools you're using the technologies is it like tensorflow is it pi torch is it just basic I mean rod yeah numpy or yeah deep learning whatever you can say yeah I don't want to you don't have to give away everything but no I'm glad to share it's been an adventure actually we started off with neural networks in the early days we had we had lots of different models going on one of the most neural networks we had we didn't have much data but surprisingly the neural networks still are performed the other models that we had but currently we have settled down on probabilistic graphical models Wow yeah it makes it really easy to recommend the next row clinics next question Excel and just like think beyond the scope of the parameters that you've trained on they are the ones that work best for us because at the end of the day we need to be able to explain why would think malaria and not dengue right and the best way to express that is through probabilities yes right and yeah and just to touch one on the stack that we are using so we have different services on our back end we have the pediatric service we have the circle cancer service we have the LCD service etc and each one of these there's a triage model that determines which specialist algorithm you see yeah and yeah so different parts use different algorithms for the most part we're using pgms probabilistic graphical models but for things like cervical cancer or things where we only had one type of data we're using different types of models like isolation isolation forests and one class support vector machines Wow okay so you're trying out different models and seeing the results and comparing them yes yes and we have different models for different age groups genders what do you believe the future of AI looks like in Tanzania is there a lot of interest in it or yes there's lots of interested in and I'm very biased with this answer so that's just a little disclaimer there but I'm glad a lot of people aren't considering Tanzania and a lot of the African countries it gives us a chance to really put our heads down and get to work and I think this is where I actually honestly believe this is a technology that will either widen the gap between those who have and those who don't or close the gap or we can go a different direction and I'm really investing in this technology even locally but yeah i think i think the ecosystem here is thriving it is really really exciting after meeting Ali I had the chance to meet his co-founder Megan Allen originally from the United States Megan had been living in Tanzania for several years which definitely made for an interesting conversation so Megan tell me a bit about yourself where are you from I'm from Arizona originally the United States yes and did you spend did you spend a lot of time there or did you like immediately come to tons on yeah yeah I know as I was born and raised in Arizona I actually went to ASU go temples and then I let see I started traveling when I was 19 I lived in India for a while doing work in health care and education which is what I do a lot of now and then I moved to Tanzania in 2015 you moved to Tanzania yeah okay so you live here yes Wow it's not you're not just visiting no have you been here then for four years yep yep that's incredible yeah Wow you don't see a lot of Americans living abroad in East Africa yeah generally that's amazing so um why Tanzania why did you choose tenzen yeah so I actually came here originally for two reasons the first one was I became the director of a nonprofit organization that does health education um so I moved here to take over that position and really grow the organization here and I was also finishing my master's degree and my research is in HIV and I wanted to do my research in northern Tanzania so those were the main reasons I was originally here and then I just stayed because they love it so much how did you come to dr. Elsa yeah so I met Olly um not long after I moved here and he was working on AI and technology and we were both interested in health and so it was just like a perfect match and we started building doctor also right after that he had already actually had the idea to build something in healthcare to figure out how we could utilize AI to make a difference for people around their health and that's something I was interested in so we delve in and we've been working on it since we eventually set our good-byes to the doctor Elsa team and took off to the Serengeti desert to interview Park Rangers using drones to detect poachers at night in order to save the lives of animals the idea is that these drones can cover large areas very fast and use computer vision technology to automatically detect the presence of humans in the Serengeti things seem to be going well we did meet some Rangers but in the middle of the trip something totally unexpected happened so in the middle of the Serengeti desert I ended up experiencing a panic attack it was this gripping terrible feeling it was in my throat I felt like I was going to die and I had to leave so we didn't end up getting the footage that we wanted from darussalam the next city that we were going to visit but I still want to show you the developer that's working there in India saya on his he's working on a chatbot so I'm gonna show you what he's working on after this but before I do that I want to just stop for a second and acknowledge the fact that I've learned a lot these past few weeks you know I've made a lot of mistakes some of them very public for example the research paper that I copied and pasted parts from from other people without properly attributing first of all I just want to say I'm sorry I definitely could have done better and I'm gonna definitely I'm striving to learn from that mistake and every mistake that I made so far and just remind myself that I'm gonna inspire I think I in general from reading lots and lots of positive feedback on YouTube like sir Rogers nai God and just like really internalizing that I started to overstep my balance you know I'm not a researcher I'm a youtuber and I just overstepped my balance and you know I think in general my ego just started to grow a little too big and I thought I can do everything I can run a YouTube channel I can do this and that and it's just not true so definitely a huge learning experience for me I had to definitely just you know just really think about what my decisions were and what my decisions are going to be in the future you know how what are my next steps here you know and I think the general consensus that I've come to myself is that I want to continue to inspire people and I want to be treating myself and I don't want to let myself or you down or anybody down and I just want to continue to create more experts in the world by inspiring people however I can through YouTube videos so all that being said it's been a huge learning experience and now I would like to show you the story of each and gauzy the sexual health chatbot run by an incredible person in yo saya and his team in dar Salam here we go yes aiya ottoman a developer living in Daraa Salam is the founder of Yungas II a chatbot that provides free sexual and reproductive health education to young girls because there is a stigma around discussing sexual health topics in that region many girls feel more comfortable asking their questions to a chatbot than to a human out of fear of judgement early signs show that it's already decreasing unplanned pregnancies in the region and the Esaias team is hard at work improving it every day I believe that the future of health care lies in the hands of ordinary people who are motivated to do extraordinary things people like Ali and Megan and yo saya they saw problems in their community and learned the necessary software skills to be able to provide solutions for people and guess what you can too the algorithms datasets and educational resources you need are all freely available on the internet you just have to get started learning I've got links for you in the video description Kulu Kulu asleep a song kama Talent oh yes the I won't lose [Music]