Get ready for the future of Healthcare: Digital Health
So good morning, everyone. And thank you for joining us on a Sunday morning. I am retro. I'm a part of the sales team. At talentsprint. I had sales for all the executive programs at talentsprint. Today, we are joined by Professor Fernando yellow Bertie, who is the program coordinator for this digital health and imaging program that we have launched with ISC Professor funding the university comes with a whole set lot of experience in this field he has been associated with ISC since the year 2008 in various administrative and teaching roles. Before that, he has done his postdoctoral research at Washington University School of Medicine. He completed his PhD in biomedical and medical engineering from Dartmouth College. He holds an MSc in engineering from ISC along with MSc in physics, so What we are going to do today is that we will spend some time to try and understand about the digital health program that we have launched. With ISC. We will also try and understand how this program will help professionals in their career, the sector, how this is a booming sector. We will also take up questions as a part of this webinar, why the questions that are being there is a dedicated q&a section at the end of the webinar, I would request you to wait for it the professor is going to be presenting he has some slides that he wants to share with us. And then we can take up q&a. So I would request all of you to hold on to your questions for the q&a part, you can raise your hands we will let you come online and ask your questions or you can send your questions in the form of a chat and we are going to be answering that as well. So these are some of the ground rules today. So Professor over to you, we can see your presentation. line we can start now.
Okay, thank you all for joining on Sunday morning. What I would like to do is that I have a little bit of agenda, what I would like to specify is give a little bit of introduction about what is objectives of the course. And a little bit talk about Delaware is a digital health and imaging is currently located. And also talk about modules in terms of syllabus schedule, and also talk about a little bit about target participants, what do we expect from you and what can they expect from us. And then finally, like more data through said, we will have an interactive session in terms of q&a. So I'll just get started off with so if something is not clear, you know, you can always put it in the chat and then I would like to answer it at the end of the session. If you have any suggestions or anything like that comments, please feel free to add it to chat. So, we can do that in the interactive session. So, so, mainly talking about the objectives of the course, this is how we have put it, we thought that first we will put out main object as a learn. Learning Objective the learning objective is how digital has transformed the healthcare and wellness. And we are going to enable the participants who are part of this course to explain the key issues and trends in the digital health particularly our focus is going to be on what is called as developing or emerging countries. And hopefully the participants will be able to determine the benefits and shortcomings of the current technologies, especially those which are available in digital health and imaging currently, and, and with with that, by the end of this course, we the participants will be able to know how how to analyze the signals or the images that are there common in the healthcare and perform task at hand using AI and machine learning or deep learning techniques. So this is what is going to be the main The objectives of the course. But what we would like to do is that we would like to enable people to know where is the information is available, and to give a tip of the iceberg or just give introduction to these topics in such a bad way to get to get started on these very, very emerging topics as well. So, today, when we talk about digital health, we moved from a traditional healthcare which is primarily driven by the physicians or healthcare providers to something we call it as consumer. So digital health is mostly about the consumer who is like a patient are not really a patient could be normal individual as well. And we are talking about health care which is going to be driven by an individual rather than by healthcare providers. So in this space, there are many things that can happen right. So you can think about starting from like in a smartphone applications, consumer variables and you can talk about interventions you can talk about continuous monitoring. So, there are many, many digital health tools which are currently available and they are, they are still emerging, they are still being developed, there are still a lot of things that are happening currently. So, if you call it as tools, there are some tools which we have, which you can see on the right side of your screen, you have this continuous glucose monitoring, you know, something like you know, fitness, Fitbit is one example there, there are electronic patient records which are there, there are consumer applications which you can really monitor everything. And then there are these collaboration platforms, right so you can think about like a social media being one collaborative platform and the other could be like, you know, something which we can post your data and get an opinion on this to be telemedicine where you are chatting with another person who is an expert on a health care provider, and then are with this all the platforms and as well as tools, what we would like to do is that come with an investigation. So that means that we are primarily talking about analytics based on this data, maybe some amateur visualization. So this is where the majority of the district has technology developers working, both in terms of developing the tools and platforms, and as well as performing the investigations in terms of analytics and visualizations. Of course, this doesn't distinguish from healthcare providers to the consumers, hopefully, we are going to enable the consumer to have all this in one single place. So then, in the imaging space, whenever we talk about imaging, we talk about mainly three things. We talk about visualization of a lesion or a disease, and there isn't any logical interpretation because of the visualization. And then primarily, the third one is an older ondina diagnosis. So you can think about like ruling in a cancer or turning out cancer. So there is some acquisition into the imaging, so which is on the far left side here by the scanners. And there are these physicians are the clinicians of radiologists who are the primary users. And the main majority of the imaging technology developers are coming between. So then that means that they take their position data to something which is which can be easily applicable. So, mainly we are transforming the data which is kind of done by the scanners to move into something like an application domain. And in between this is where we come we talk about competing and analysis is being one of the majority of the tasks here. And most of the imaging technology developers work in this space. So, this could include as simple as storage of the data to on the way till advanced image processing applications, artificial intelligence applications. So, that is about roughly about where do we stand so that's why we have put both digital health and imaging as one single course. Because most of it is technology development is what we are planning to do. And again I told you right this whenever we talk about digital health that we talk about consumer products, so, we are no longer talking about in the space of healthcare, we are knocking in space. So because of electronics and the Consumer Electronics Show, which is the one of the biggest shows of the Consumer Electronics has a dedicated digital health summit. And, and so for example, for this year, it has happened between January 7 and 10th.
And if you really look at it, remember we talked about health care, you know, digital health is one of the biggest one which is out there. And this connects to both data with this connects generation this connects to big data. This connects to technology as well as technology development as well. Of course, healthcare is at the at the center of things. But if you look at it, what is happening currently in terms of discuss topics, they still have major Be able to discuss topics are in terms of artificial intelligence, which is about 42% of the topics. And there is a little bit of application developments which is about 27%. And predominantly in the recent ones there is a blockchain and there is a little bit of augmented reality health insurance and the color industry, which is mainly to do with nutrition and things like that. So, these are all the primary discussed topics in the in the digital health domain. And that's why if you look at the course, at least they we have two modules dedicated to here. And the rest of them are in the application specific or application. Of course, when we when we talk about modules and, and things like that it become little more clear for you. And one of the majority of the thing which many people don't realize if you are told you are planning to jump into this healthcare, and as a service industry, if you really look at it, it requires a lot more continuous learning I mean, this is just giving you a chart of you know, what is the, you know, timing, which is required roughly compared to other service sectors. So if we look at health and medical service requires majority of the training, which is roughly about 16% of the time has to be dedicated to time. So, this is roughly in terms of time. So, if you think that you have 60 hours in a week, which you are applying to work out 40 hours in a week, you got to spend about roughly one sixth of that time. That means that almost one day in a week, you have to work on getting a new training a new skill set. So, now, let me jump into what are the models which are currently we are planning to do, as I mentioned that distribution is a big topic variables are one of the biggest topics in the digital health, because most of the data comes from the variable. So we have faster we give just introduction to digital health. I'll get into little more details in the next few slides. Again, we talk about wearable devices, we call it a physiological signal processing physiological signals how they acquire and things like that. These both form the distal health modules. And as I told Tao as a as you are already seen most of the topics of the digital health are involving around artificial intelligence. So we can move away from learning machine learning, and then talk about a little bit about deep learning additional, and then get into the imaging and that vision. So hopefully given more into the code situation, when we come to the module five, we are going to have a contact, you can have a campus visit. And then hopefully, it's all going to be hands on. By the way, all of these except the modular one, the rest of them all require you to run some demos. So we will give you something like you know, hands on demonstrations for you to try out by analyzing the data, which is out there. So, just to get into a little more, you know, granularity in terms of giving you what are the other more So, what are the topics which is covered under each model? So, for example, if you look at the first module, which is about digital health, we talk about the need for it, what are the case studies? Talk about a little bit of the basics. You know, as I told you what applications m health is mobile health is one year is another, what is the current impact in terms of implementing what are the current guidelines which we have current standards in terms of HL seven, integrating the whole healthcare enterprises, we talk about something called as vendor neutral archives. And we also have to talk about open source data innovation. And more importantly, we have any digital health requires some IT infrastructure. So, we are also going to talk about IoT and cloud computing in terms of the first morning in terms of the second module, we are going to start off on the variable devices in terms of physiological signal processing. So, what are the basics for it? What are the different sensors which are currently there? What is instrumentation aspects of it? What is the data communications aspects of it. And today audio has become a very big unanswered for this, most of the natural language processing happens in that. So stop talking about speech and audio signal processing, starting with signal capture data pre processing, look at feature modeling.
And then we jump into the machine learning which is the aspect, we talk about data model support vector machines, which are traditional machine learning ones generating models in terms of expectation maximization and latent variables. And then we get into doing little more hands on in terms of deep learning in digital health. First, we talk about machine learning for physiological signal processing. Mainly we talk about time series, because most of the data and the physiological signals is mostly a time series data. And then I talked about a little bit about the basics of the deep learning in terms of CNN and things like that, and give one or two examples of deep learning for physiological signal processing and natural language processing in digital health. We are also To cover that, and the last module is about the imaging, we talk about medical imaging modalities, what are the typical analysis which we do? What are the typical challenges? We talked about applying those to the deep learning topics in terms of segmentation, you know, convolution neural networks optimization, and what are the current deep learning models that are available in terms of translation, starting from unit Alex net, VGG. net and our nets. So these are really giving you granularity in terms of water not be covered as a topics in the course. And so in terms of instructors, we have a very good instructor set. And I'm going to mainly handle the module one and parts of module two and five jaiprakash who is asked in person biomedical instrumentation, he's going to primarily cover the module two, which is about wearable devices instrumentation aspects of a person aspect. Who is a person artificial intelligence property science department. He's going to handle primarily the basics of machine learning and come back and help us in terms of model five as well in terms of deep learning models, sit down was is a is again by deep learning expert he works on mainly on audio and speech processing. He has also has experience with IBM Watson, which is the biggest Deep Learning Network which is out there, he will handle parts of models too. And for Pankaj Babu will handle parts of model five, which is part of the computer vision because most of our models are derived from computer vision. You're going to show again, he's a piece of data science. He's going to handle part of the model one in terms of IT infrastructure is cloud computing and IoT. But we are going to start hopefully on August 23, and all the classes are going to be on the weekends, Saturdays and Sundays and each day we are not going to burn you for more than three hours. You can plan to attend the online sessions. Or you can also look at look at the videos which are being posted after the recordings, you can also look at the recorded videos, our plan is to finish by December 20. Assuming that the coordinating the campus is going to be open, either we will be able to access the studio I think like that. And most of the demos except the end where we are going to do a little bit of hand holding, the rest of them are all going to be install and run kind of demos. So that means that will give you certain you know, combine it developed models, we just want to run and come about and get a hang about analyzing the data and then get a hands on experience in terms of looking at what is outcome of that demo sees a developed language. Now again, so in terms of participate, for what we have targeted is mainly a tech developers. So that means that you should be willing to make your hands dirty. And then will be willing to make the technology in terms of writing algorithms, writing the codes and things like that. Even though you may not be really doing coding as such in this course, but we expect that after the finishing of the course, you are going to be in the technology. And we expect that you know, the basics of the programming and you know, how to program things, especially like things like, you know, elastic stack, you know, kind of what an industry standard in terms of time series data is analyzed, analyzing the time series data synchrony, like Python, and TensorFlow, things like that. And of course, we require you to be passionate about issue technologies for the healthcare. I mean, obviously, if you aren't all you are spending time here and there as well as your money. We would like you have to be passionate about that. And so there's that I would stop by I'm very happy to take any questions. I have my email id here. If you have any questions, you can always email me. And so I'll move to the interactive session.
So thank you, Professor, thank you so much for the very informative presentation. The next question comes from Mary Joseph. I believe she's already applied for the program. She doesn't have experience in Python, but she knows programming. So her must be able to understand the course she's an MTech in biotechnology.
Now, I think if you know programming, you will not have any issues with the course because most of it requires a simple installation of some kind of framework like pi torch or TensorFlow. And you just need to run these programs are called there are many, many online sources which are available, if you want to really code it from the basics are coded from the scratch. So even though we typically all these demos which we are going to do have a pre trained models, which is already trained, and all you need to do is running these models, but you can actually, if you have the access to the GPU, Use and if you have enough compute power, you can always train it from the scratch. And those are very easy.
That's a very interesting observation professor. Moving on. There's a question from Syed shootie. She asks, How is this course different from AI in health science? Is this more focused or a specialized course?
Okay, so that's again, a big, you know, and I That's a big question. I think you're putting me into the corner. But having said that, as I mentioned, right, so, yeah, you can help scientists one topic. When I say but I look at healthcare, it has many puzzles to it. Let me just go back and then show you why. So, can anybody everybody see this slide? If you ever see yes. Yeah. So you know, you look at it. You know, there is a health stack. There is a data there is innovation, there is big data, there is tech and technology. And yeah is only one piece of that puzzle and then in the sense that okay my drive majority of the things currently that are going on, but there are many other things which which is not really AI based. So, for example, like we talk about things like you know, still believe the comfortable in making a diagnosis by machine will you be more comfortable if the machine diagnosis is verified by a condition? So, most of the people will answer I can tell you that 98% of the people will say that second one is what they prefer,
we will put up a team of mentors who can help individuals people see we are going to have a very small cohort of 50 people in this program. And we will have mentors who will be available on a one on one basis to help on a one on one mentoring slot booking. Right so if for example, 80% of it you're able to I'm getting stuck in the last 20% in terms of installing the pytorch for example, our team will be there to help you out and complete it so that you don't miss out on the learning that is necessary by completing that exercise
Watch the entire interview here https://youtube.videoken.com/embed/I3nmx-XXoZ8