IISc Digital Health Program
I mean what do you mean
just check with Professor no problem
Hello Ramesh? Yeah. Good evening. How are you? I'm doing fine. How are you doing? All good, all good. zephyrhills I'm really sorry. There was some disturbances, so could not share the link at the earliest. Sure not a problem.
We'll just start in five minutes. I'm just waiting for Professor metka. Once he logins we will start. Sure yeah. Thank you. Thank you.
Can you hear me now? Yes, Professor we can How are you? Good. Good evening.
So now we can get started.
Yes. Let me just share my screen and then we'll get started.
It's my screen visible.
Good evening, everyone. I'm really sorry there was a delay of 510 minutes because of some technical glitch. I hope my voice is audible to everyone. So I welcome you all to this information session where we are going to speak more about digital health and imaging program, which is being offered by Indian Institute of Science, and nse talentsprint. I hope everyone is doing well and they are safe at home. Allow me to introduce myself. My name is Ramesh Monti, and I take care of this program and I'm the program manager. This will be my third cohort with the Indian Institute of Science.
Now, today's agenda like similarly, we had a session or a webinar one month back earlier. So today we are going to discuss on the same topics I mean, what are the object of the schools or the curriculum format duration, the prerequisite and all this stuff. We also have invited Professor Ambedkar and Professor jaiprakash from IAC to share more insight about this program. The last segment, we will allow you to have all your questions to be get clarified. So I request all of you to post your questions only on the chat box.
Before starting I would like to take this opportunity and speak about who we are and what we do on behalf of talentsprint. Now talentsprint is the National Stock Exchange group company top academic institutions like iisc, Bangalore, IIT Hyderabad, triple IIT Hyderabad, I am Calcutta, IIT Kanpur and recently we have also IIT Madras have partnered with us to provide deep tech programs for working professionals. We have multiple programs of these Institute's and for more information about these programs and Institute's I request you all to log into wwe.talentsprint.com
now, allow me to also introduce you professor Ambedkar and Professor jaiprakash from IHS
See, Professor Meeker is an associate professor at Department of Computer Science and automation. He is an AI sc and IIT Madras alumni. On the other hand, Professor jaiprakash is a professor in computational and data science at IAC. He has held postdoctoral positions at universities across Europe and worked in industries before taking active interest in this academies. Excuse me. Rubbish.
Rubbish. Your PPT seems to be frozen. Can you please? prompt?
Is it fine now?
Yeah, now it's looking if you can make it. Yeah. There's a full screen now. No, it's not full screen.
Okay, I'm okay.
Okay, this kinetic analysis is
It's okay. I wish let's send to me. Okay, so
I'm sorry, Professor for the dissidents. I welcome you both. And thank you for taking your valuable time for this session. Before I asked Professor Baker and Professor jaiprakash. To start with this presentation, let me also tell you about the last cohort statistics. This will be our third cohort. And for the last two cohort, we had on average of 65 students per batch, and 60% of them had more than 10 years of work experience from different domains of healthcare sector. More than 50% were into senior executive roles. Or you can say leadership roles. Out of this 38% of the students were co founders or they have their own startup. And last to code, we also had around 20%, of human participation for this program.
Professor Ambika and Professor jaiprakash I would request you to start with your presentation. And let me know if you need any help on this. And I'll be stopping my screen right now.
So can we start?
We can start Yeah, I'll share my screen.
Okay, everybody can see my screen. Yeah. Yes, perfect.
Good evening, everybody. My name is a maker and our faculty in computer science department.
And welcome to this session and today, myself and process jaiprakash will
give information some information about this, you know, digital health and imaging
actually, so I come from the machine learning. My background is computer science and pressor jaiprakash actually, is an expert in this area. And I'll hand out to Professor Jay Prakash
introduce this program and
get in there after some time,
just jaiprakash so you can get stuck.
welcome all and I'm hoping that everyone is safe and sound. Okay, so, the agenda for today's presentation is that I will introduce to the intended use objectives of this course. And then we'll lay basically introduce the different aspect technical aspects of digital health and imaging and then we will move on to the different modules and and all the faculties in Word what is the schedule and what are the prerequisites required for this particular course. And lastly, what how do we evaluate and out of the last last we now also have interaction impact position where you can ask the question and answers and we will clear the doubts. Okay. Next slide.
Yeah, so what are the objectives of this course. So,
for the objectives of this course, is to know
On how digital health transforms healthcare, and the wellness of
So, out there, how digital health will transform healthcare and wellness. And next, we would like we want the participants to be unable to explain the key issues or trends in digital health, particularly focusing on the developed countries because that's very a lot of work to be done and that will help you to
help you in your carriers and so on. And the participants will be should be able to determine the shortcomings of the current technologies in digital health and imaging. So that's what we want the participants to learn and at the end of the course, participation will be able to analyze the signals our images that are common in the healthcare and perform tasks using AI, ml or other mathematical tools.
So, what is digital health digital health is a is a vast area it covers as you can see here it so it covers a lot of topics, first starting from consumer health, consumer
wearables to the different mobile apps that is that is developed for for using these wearable devices and also telemedicine and analyzing the patient health record and trials all these come under the umbrella of digital health Okay, so on digital health is moving. So, the that's also the treatment regime is moving from hospitals to more and becoming more patient centered. So, you can see here that
you will have a patient and the patient will have access to the different digital health tools. So, you can see that you have access to countries people's monitor ECG eg and also something like altimeters, blood pressure monitors eye are accelerometers and so on right. So, all these variables will be will be discussed at the disposal of the patient and what as digital health technology developers our job would be to analyze all the signals that we are getting from the different
devices and give some kind of
some kind of
some kind of
overcomes okay and sure very good visualization of outcomes in the sense of showing not using some kind of a bar graph or all those kinds of tools, okay, but that is what is the goal of digital health technology developers and in between this you also have collaborative platforms wherein you can see that there's something steady medicine in telemedicine, what happens is once you analyze this data that will be directly sent to the doctor and then the doctor will try to interact with the patient and provide some kind of guidance and there is also this collaborative platform where the the all the people who are work as a community and there will be a community goal established for example, if you think about fitness, right. So now there are websites where where there is fitness goal, which is where you achieve targets based on as a community rather than individual. Okay, in that case, what happens is
your, everyone works together to achieve that goal rather than having having an individual. So that's what it's called, this is called as the social or collaborative platforms which are available which are made available these days. Yeah. Next slide.
So, what is digital imaging? So digital imaging basically has three steps okay. The first step is to visualize lesion
and or disease. And once you once you visualize this lesion, that that particular
images will be shown to radiologists, so, they will do some kind of radiological interpretation after radiology interpretation they will either rule in or rule or rule or diagnostics, okay. So, basically, you have three three different aspects here one is a standard one one is a scanner at the scanner stage where you do the acquisition of the data. And this is where you have all the instrumentation and physicists who are willing to work and acquire all these data, okay. And at the other extreme, you have users who are basically doctors who get these images and I will properly visualization and they try to analyze, they try to rule in or rule out diagnosis based on these images. In between is where you get this computation scientists where the use where they use acquire data, and then do they use computing they do completing an analysis and
I'll show the data to the to the radiologist in a meaningful manner. And this is the place where you have all the digital imaging technology developers or people
They have great role in terms of using image processing or ms reconstruction or machine ml books.
So, if so, if you're if you are in the working area of digital health right, so, there is a very big show called Consumer Electronics Show which happens every year this year also happened in Japan virtually. And if you see right, so, digital health here is linked with all all different technologies starting from big data or innovation and everything. So, healthcare is a bigger umbrella and digital health is related to all these different concepts, okay. And if you see the growth right in the area of digital health, it is currently in 2018 it was about 144 point 2 billion USD and it was expected to grow to 230 6 billion USD by the end of 2020 which actually comes comes out to have a component and your annual growth rate of about 30% which is a very big growth rate for any industry. So that makes it very attractive to work in the in this digital health space, okay. And if you see what are the topics that I was discussing, this Consumer Electronics Show, you can see that artificial intelligence was about 42% of topics that was discovered, that was discussed in the Consumer Electronics Show. And after elopement was about 30%. And blockchain is about 20%, and so on. So we are not going to cover blockchain in this particular course, but the other topics we will, we'll touch on.
So, what is the what is the biggest need need in a when you're when you're working in a domain, so the biggest need is to learn continuous, okay, and this is not like any other any other areas. So, if you see the different areas, here, you can see that health and medical services request 16% of learning, okay, compared to other other areas where even if you see something like education, right, that's only 6% so human faculties need to learn very less and get upgraded very less compared to people working in health or medical services. And other areas. Like
if you can think of business services or government or any other area physical bank banking and finance which are which has a lot of money, you won't even that has less learning compared to health and medical services. So basically, if you want to work in this area, you need to update yourself constantly. And if you're not updating users constantly, it's going to be difficult to sustain in this particular. So, this is very important important point if you want to continue to work on a test.
So, how So, what we have done is we have put up this digital health and imaging course and it has been divided into five modules. So, you have model images, digital health, where we introduced interior to the digital health concept and your model to where you have wearable devices and physiological signal processing these so these to form the digital health basic modules, and visual and then Imaging Lab all the five modules even the capstone project, okay, so module three will deal with machine learning basics for real world a module four is deep learning in digital health model face deep learning and imaging ambition. And then finally, you'll have capstone project, which is you only need to do a project
project from a list of projects that has been put up, you can you can also define your own project
with some constraints that some might be isolated constraints. So, this this is what is this how the syllabus is defined. And one more thing one more important thing is in each of these modules, you will be given some kind of small programming assignments which you can do on your own and that will help you to understand the content much better. Next,
oh, so, in the module one, the residual health these are the topics that will be covered Firstly, what is the need for digital health and the different case studies that will be introduced in the digital health space? And next will be the basics of digital health and what is mobile health and so on. And how did digital health impact the current current
current state, okay, so And
next is informatics where you will be learning about edges and standards and how to how to have these in how to do this interoperate on operability of the data. And know you will also will learn learning about vendor neutral arkose and open source data generation and the opportunities for this opportunities that are available when you when you're working with open source and open data and so on. And lastly, we'll also be thought about IT infrastructure that pertains and cloud computing and how these have
Are these health digital health
and how these are well revolutionized digital health and the analytics on domain.
And the next module is basically variable devices and physiological signal processing. In this particular thing, we will introduce you to signal processing, where you will be taught about sampling basic filters, and dissemination interpolation, short time Fourier transform wavelets and these concepts and then we will move on to the concepts of physiology. And in physiology, we'll talk about ECG signal acquisition and basic physio, physiology of a heart like, what what is it like to their heart? And how do you play the chest while acquiring ECG data and so on? And
how do you read ECG data, and next we'll move on to eg signal generation and some kind of Asian ECG data processing. And after there's really more than two wearable sensors, well, we'll talk about glucose sensing devices, accelerometer, wearable, ECG punnamada gram and so on. And nowadays speech and audio signal processing has become very important. So, this this topic will also be covered where we start, where we will start from all the way from signal capture to data pre processing and modeling the feature feature vectors and so on in the speech and audio signal processing.
So, I'll give it on to maybe
Yep, thank you.
So, like to know, what we have seen in that in module one and module two, you can see that there like,
there is a huge opportunity if you know,
to to understand the problems, there's a lot of you know, data is being generated, right like, So, if you look at the hospital records, you know, various governmental initiatives and sensors. And you will if you look at one particular patient, like you know, a lot of tests
and tests and prescriptions and various amount of various kinds of data that is being generated, right. So, how do we make sense of all these things.
So, that is where machine learning
comes into picture, right. And so, now, machine learning and AI, like, became a very big buzzword, you know, if you try to learn and it's a huge ocean there and, you know, sometimes I feel that the most difficult part about machine learning is that you know, where do you want to start and the way you want to end right, and
so, what we have done is that we have put up a program in such a way that you know, we you can get some kind of a unified view
of the subject, right. So,
she did the two ways of understanding machine learning the machine learning like there are 100 methods like or causal methods like a set of 100 Auto the method or or you can understand as a field where like, there are some foundational principles and like some some foundational paradigms and based on which everything is built upon right. So, that way, you are if you learn two methods, you understand them you understand them as something related in a certain way rather than like you know, this is like a metal one and not a two like that, right. So, particularly module three if you look at you know, it's also requires some basic mathematics right,
anything that you want to formalize right let's if you want to think in a formal way, if you want to say something in a concrete way, right, it requires some bit of mathematics right. So, for machine learning, it requires that you know, you should
some amount of
probability and linear algebra these kind of things that are required and in the first lecture, so, we are going to teach these things right. And also what is data and model like, you know, machine learning applications.
then, we talk about
introduction to real world signals, text speech and image and feature extraction methods and
so, so these kind of topics and also we talk about something as learning as well.
So, what is right? So, this is the very fundamental way of looking at division learn right. So, everything involves the optimization, right.
So, you have some kind of data, and you design what you want to optimize, and then using the data and want to learn you can do it in a
with a probabilistic model and without probabilistic model. So, all these things we are going to cover, right. And
then, you know, nowadays, you cannot talk about
without talking about deep learning, so, basically what is deep learning like these servers, what, again,
deep learning deep neural networks was there in 80s. And there is one algorithm called back propagation algorithm. And still we use the same thing, what that has changed is that, you know, we have access to
access to fast computers, you know, so that we can run these models and get
very good results. So, one can actually say that deep learning has revolutionized a lot of fields, right.
Our accuracies have gone up. And there are some things which some of the tasks which we could not do before we are able to do, right. And so, for example, and give you an example. So, if you look at the medical records, right, like, see, there is an image and there is also some kind of a text, right, so, these are all we call it modalities. Right, so, so data comes in a different kind of modalities, right? So, one of the amazing thing about this deepening is that you can deal with all these modalities in a unified way. Right?
Say a simple, non medical example, is that
speech, right? It didn't watch the language. And if you're looking at the text, right? And so how do you describe the sentiment in a text, like,
if somebody is actually speaking, if you involve the speech and the text, we're going to actually capture the right. So medical records, the one of the important examples of the,
you know, dealing the
way several modalities come into picture, and deep learning place, and very important. So what we'll do is that it's not some people, what they do is that, you know, they'll download code, like CNN or something, they'll try to learn these things. And
though it gives you a short, immediate feeling that you're you have achieved something by actually start currency training these models, if you really want to solve your problem, right, you have a data, it's not that easy. Unless otherwise you understand what was training these methods is not going to be that's what we are going to start off by basically, we are going to perform the foundational aspects.
then we'll particularly have module five will concentrate on deep
imaging and vision. And
so, we'll talk about
some particular models and
some amazing physics based methods and need for deep learning and neural imaging and this kind of thing. So, we will cover there are some important pre trained neural networks that are called the Alex Nash VGG and Google lead and respect and and these kind of methods they are going to give you a hands on training right. So for example, if you look at me segmentation, right. So which is a very important problem in medical imaging, and which is you do something called the unit right, this is the neural network. And what we will do is that we will give you an a hands on experience with the hands on training on these, you know, these models, right?
So, these are the instructors and professors. phanindra is an expert in medical imaging. And we'll handle module one, and module five, and procedure Prakash, and we'll have the model to myself will hand the three and four along with the procedure. We're going to pathy
and then dish mobile's
expert in computer vision, and he'll handle module five which is limited image
So prusa Yogesh simple, he is expert in data science and he handles
combining all our expertise, so we all come from different backgrounds, right. So,
so that's where this course is very unique.
So this is the core on the target participants, right? So this is targeted at technology developers, and we emphasize that, you know, programming is very wide. And I'm sure that our sprint is already
ready has started this course, see, this is the Python, I know some of you might be familiar
with the beginning
something language, which is not like other language C or c++, this is like almost
natural language, right. So, as long as you don't have that, so, we are not asking you to actually code from scratch, but at least when there is a code, you should be able to understand, like, you know, which part handles what right. So at least, so that you know, it can make you appreciate the content much better, right.
So, so, you should have, of course, you should have, you should be passionate about the digital technologies.
at least work at least three hours per day,
more than I think.
So learning is not difficult, as long as
you you spend some up disciplining and spend some some time a number of our safety, right? And not afraid of change, and they leave a hard and smooth in mind is a very important thing. What happens is that
not because of our voltages or something, it's going to happen to our students, like who are supposed to be and so as long as we started teaching, some some topic, which is new, or slightly mathematical, they switch of their mind, right? So that's where it's starting to do the age, right? Like it's just an attitude which can be changed right, like any time
I know also not that they don't take it as a burden just joy right, let's just enjoy the things
used to make the hands dirty. So, for example, in this pose, savers will give you whichever the poor of your unit or something like that, and we will say that you know, try to make some modification and see observe, so, that is very exciting, right, instead of just running the code and leave it like that. So, just make some changes to the hyper parameters and see the power
chain, then you can relate
the concepts that we have covered to help practically is there. So, we are restricting to 50 participants, because
we want to be very interactive and we want to
So, we cannot
be we cannot maintain this interaction, there are too many participants. So we have to
So, this I have already mentioned some of the prerequisites.
So, so, this is the warning like should you be worried like because
this is almost like a semester long course right. So,
some of the prerequisites that you require are the basics of signals and systems for example, you should order for a transformer transformer.
So we do we start our lectures by not that you know, we completely assume that you know all these things, we obviously start our lectures
by covering these things and giving the definitions etc. Say for example,
when I went to machine learning
I will do a bit of basics of probability, but I my lecture is not going to be on the basics of probability, right? So we will,
we'll touch upon these prerequisites, but we move on to our aim is to learn machine learning, right? Not that structured like this property. So these, see, maybe some of you might have learned all these things later. We want to learn later, but you're not going to work like that right like, So, during the lecture, we want you to appreciate what we have covered. So this way, it is better for you to have these kind of basics, though, we will anyway, touch upon these things in our lives.
And some kind of programming experiences must like
that you may not
write programs from scratch, but you should be able to understand that code, and it will be changed the board.
So if you look at the portal of the HCA and you already the bridge course on, you know, Python has already started.
so this is what so we're starting on
this course, in the first week of September, right. And this is September.
And so the beach course will be between June and July or August, right. And then be covering some of these Python basics and TensorFlow and
some of the
some bit of understanding of like, you know, CNN,
able to change these things in the TensorFlow, right.
So you can take a look at, so I think we are going to share this slides with you so that we can take a look at these slides.
Look at these links and see what you have.
for Python basics, you can take a look at this YouTube video.
posts on Python for data science, and
pytorch exposure. So particularly,
deep neural networks is good to go by. So basically, it's a Python based one.
So each of you explicitly writing the each modules, you know, the display the neural networks. So basically, all these things are already implemented, you just have to use right? So. So even the most complicated neural networks can be implemented in few lines, right, it's pytorch. To understand this, so these things you can take a look at, since we have time, we want you to really warm up and come for the course. So there'll be validation. And every module contains a simple activity in class. And will be please select the multiple choice questions.
The bigger person and for those who don't get 80 80% and will be given for task two, for every module containing Grade A plus A, B plus
each session will not be more than two hours, right max two hour or four hours in a day, like morning to us in the
two hours, right.
So at the end of all the modules,
it has to execute a project.
all instructors will propose some set of
projects and we can choose
this project, I will also make sure that you know you will have some food available to you. And then we'll have a lot of interactions and so that you know you get acquainted with problem solving.
And the three doors that we do all this discourse is not about the key program, right? It's about solving problems. Whatever, see, programming is the only tool to solve the problem.
So so you have to understand this pitch prospect character, and it's really important to appreciate this course.
So then we have so the
Hands on the most, right like all the training
will be, you know,
included in the lecture. So, like so, for example, if I did some lecture and one lecture, we dedicate to the demo slide, where we will share some code to you say for example, on the unit, right, this is the basically with your letter, we share this code to you, and the data
once we shared this, in advance will be all the details, maybe you can play with this or we can show you what things that can be done, right.
So, the teaching part is like 110 hours and under parts of mentoring and the casting side and the 20 120 years. So, this is not a regular certification program.
So, this program basically offers, see, obviously,
spend a lot of time and, you know, not only you learn things, but also give a certification, including the transcript, if you take grades or every module from Central for continuing education, that is basically the central for the Center for continuing education, cc's also conducts their own programs, but, so, these are the ones who are going to do the certification.
So, so activities in the class pursue to attend exams and complete assignments and do a capstone project. And everything gets evaluated and given marks and grades, right. So this is not, you know, you listen and forget, like, right, it's not like that everything will be get evaluated, because you're going to get a certification.
it typically, a student in I see, basically will do a three is three courses in a semester.
That is total of nine credits, right.
So, so, so this whole course, we can consider it as a one semester spending one semester in is, right. So it is equal to two credits, that means that
so we add as the rigor as regular is equals, right?
So there'll be a one to one office hours per participant, with instructors where you can ask questions, you can clear your doubts, right? So
what is the mess? So digital health and imaging laptops are waiting for home care
costs, and I think next many years to come, right. It's going to be the top most fee in look at any other industry
that they are not that they get sick. And yeah, so
more and more health services will be required. Right.
So join if you're willing to learn on continuous basis join if you wish to make change to technology, join if you're not afraid of change, join us here at angered hearts.
So, there is a lot to do. And there are a lot of not many people
in technology are required to make some change in the digital healthcare scenes, right? Particularly in India or elsewhere. Right. So
So we hope that that's where this course is very important role.
I previously so there are doctors and there are people who maintain hospitals, etc, etc. And these all these things are changing because now, you know, there's a huge amount of data, somebody has to interpret and make sense of it. Right.
And we hope that we have designed a lot of
thought has been work has been put up to create this course. We hope that it will give
Get overall training to be a technology developer in this area. Right? So, again, technology developer doesn't mean that writing the programs, right?
We, technology developer means that innovation, like you know, so
one thing is that, you will understand that a lot of people think that EA can do miracles and almost right regulate can do anything. It's not, it's not like that. Even if you if you have a huge amount of data, it has its own limitations. So you cannot start some startup saying that, okay, I'll hide some, I have some data, and I'll have hired some graduates in machine learning and try to do something, it's not that you should know what is possible and what is not possible, right. So this is where
this course will help you. Right.
So thank you. And I hope
we have given you all the information responsible, if you have anything can always contact
Thank you all for your attention.
Thank you. Thank
you, Susan, you can go.
Thanks, Professor America and professors jaiprakash. For an elaborate details about the program, I totally understand that the participants have enjoyed and understood the impact of the program and how promising the course is going to be. There's a few questions in the chat box, I will just take it one on one. And any one of you can address those questions.
So the first question.
So there's one question from Zoho core, who wants to know, how will it add to the existing job prospects of healthcare professionals with respect to clinical therapists?
Yeah, sure. So, if you're if you're looking at clinical therapist right, so, even there, you have to do this
what is that standardization and maintenance of any device right? If you look at radiation therapy, so there you need to at least in the West, you need to do you need to get up every day in the morning and do this calibration of the device or the radiation therapy devices. And nowadays, you also have this new technology which is called us
image guided radiation therapy or intensity modulated radiation therapy igrt and which are which are coming up right. So the field of therapy is actually moving towards something called as theranostics. So that is integrating therapy with diagnostics in some sense, okay. So since imaging is is one of the core core topics in this particular course, and how to how to analyze imaging data and,
and use the imaging data is one of the core concepts. So.
So in the future, you will have a lot of applications of AI in imaging in the, in the condition of image guided radiation therapy, or intensity, or igrt. And these kind of technologies, okay. And not only that, even simple things like doing clinical trials late where you discover new drugs and understanding the drug efficacy and all these
all these aspects, you need imaging, imaging, to see how the distribution of the bad distribution of the track and so on, right, so as these tools will be useful in analyzing and giving, giving you the statistical, statistical significance of the drug, and so on. So that is the benefit of this code in the context of therapy. I hope I have answered your questions.
Yeah, thanks, Professor jaiprakash. There's another question from Mr. Swaminathan that is this force more inclined towards analytics on top of data from different sources, or will it enable us to build solutions in a cloud path platform so that data can be aggregated? I have answered that in the chat.
Yeah, thanks for that. So this is another question from Dr. rajarajeshwari.
So will the curriculum include statistical calculations using software's like Minitab are in programming etc.
Okay, I can answer. This is not as a two as the
I mentioned that we are trying to solve the problem is not about how to do the calculations, right? So you do the calculation, say, for example, you're trying to optimize the system, you have some statistical model
to make sense of some data, right? Then you follow some kind of optimization problem, and we try to do that, right. So, mostly we'll be using Python, right.
So, the aim is not towards how to do the calculations by on the models, and how do we learn the systems? Right. So that's what we are going to concentrate.
On. There's another question
regarding the certification and job processor, Ramesh would separately connect with you and discuss all the other details regarding the program and the certification.
There's another question, what level of Python knowledge is required for this program? Mr. Sanjeev is asking this question.
So, at least So, this I have explicitly mentioned,
you should be able to understand some
existing code at least. So even better is better. Like, if you are able to write the code,
but you should understand the syntax and,
and should be able to run the code and you should be able to modify that code, right. So you have enough time till September. And also during the lectures, you No one can always pick us up. Python is one of the most easiest, more than two, two language to learn. So there's nothing to worry about learning Python, like anyone can learn, right? Is this. So? That's what we expect? So so my short answer is that you should be able to understand the code.
You know, what is happening, like existing code, at least?
Right, thank you, Professor ambetter. There's a question from son case, he wants to know that he wants to work to develop new medical hardware devices. So just wants to know whether this is the right course for him.
So thank you. So in this course, you will be taught about the physics of acquisition, right? For example, if you think about something like glucometer, right, so how does glucometer work? And what is the physics behind it, but per se, we will not be teaching about the electronics or medical electronics and those aspects, because that will take a course in itself. So how to devise circuitry and all those stuff, right. So that will be a course in itself. So we're not handling the circuitry, but, but in general, we are handling the physics of say all the devices like pulse pulse oximeter, ECG, G, and giving a bit basic explanation of how this How does the sensor work, but we don't really give hands on training on sensor building or sensor characterization, all the stuff, okay. And secondly, what are the signals you're acquiring from these sensors? Right. So, we will explain you about how this how to analyze that data and how to for example, how to Deena is the data how to
extract meaningful information from the signals right. So, these are the aspects that will be covered in this course.
I hope I answered your question.
If it's good enough, Professor jaiprakash there are certain other questions which we accumulated professors is one of them is about you know, how will How will this program help the participants to translate what they learn in the class to their job. So, if anyone of you can help ensure throw some light on this
Okay, so So, so now, as I already told it's healthcare is a very wide domain where you have various people starting to work from the area starting from doctors to all the way till software developers are experts in in pharma, regulatory and all these stuff right. So, everywhere you have you have technology nowadays. So basically, what how is the database if you think about some, if you think about
industry professionals, right. So there what happens is,
nowadays, everything is is becoming digital. So you take healthcare sector or our oil sector in any sector right now.
So, the word digital is the key. So, in that sense
any technology so, you take you take sensory technologies are imaging technologies right? So, you need to understand the concepts of AI in much more detail. So, it's not as simple as just downloading your code and running the code right. So, you need to understand the intricacies of why the code is working and what is the problem behind this and so on. So, once you understand all this, it will become very easy for you to translate to industry, okay. And you have a lot of a lot of companies ready you'll see any company which is in the healthcare space right, like you have a Medtronic or GE Philips, Siemens, any of them right, all of them are interested in this particular digital digitalization space and how can how can you extract meaningful information? How can you How can you create these protocols such that it will reduce the burden on clinicians and so on right. So, that is the industry path. So, now, if we talk about specifically about doctors or academics, right, so, there The question is more about if you will be enabled to develop your own projects, okay, and which can help you to treat the patients much better.
Okay, on your regular day to day to day affair. So in for them it is it will become much more easier to treat patient and do quality care for the patients, okay. In that sense, it will help doctors a lot. Yeah. So, this is what I would say in in these two arena similarly for pharma, pharma companies, right? So if you want to go into something, you know, what is or something like
other other big pharma companies like some of you and so on, so that it's more about how to use these imaging tools to automate the clinical protocols, right. So in the context of drug delivery, so how can you use this small animal imaging technologies and how can you use AI to reduce the number of mice involved and so on, for that discovery, operations? And also after that, how can you reduce the number of
number of drugs which are not really working? Right? So these are the different aspects which you can understand much more, which you can contribute to when you're working in the pharma industry? Yeah.
So there's one question from Kellyanne again, that how this course will be helpful for MBA, like the professionals who have done MBA in healthcare and hospital administration. Yeah. So if you're looking at an healthcare MBA in healthcare and hospital administration needs, so, in that case, what happens is you need to maintain this patient records right. And nowadays, what's happening is most of the countries are moving towards bags and and those and
digital health records data and so on. So, the reason that say that we can more and more important is many scenarios, we require law, the patient history, okay, the patient history is very important. And given the today's context, what's happening is most of the patients or at least in the Indian context, they might not really know
more about their health health situation right. A simple example being if you think about it, if if a person is
having character levels very high, okay. So, now, we should not you should not be given antibiotics, right, this is a very simple example. But if the patient is not able to tell this to the doctor,
then it's going to lead a lot of complications, right, if they if they give them antibiotics, the keratins will shoot up and that will be a problematic for the patient. So, simple, choice of drag is also very, is also very important. So hence what happened, what's happening is now every everything is is going is becoming more and more digital and maintaining these records is become very important. So in that sense, when you're when you're maintaining these records in the hospital, right. So, all these all these data as has to be analyzed by the by the hospital administration and he will be the first point of contact right. So, he will be he supposed to extract this data and according to you that to the doctors and they will train them and so on. So, in that context, this this course might be helpful in understanding the entire protocol starting from So in the first module, right, where you have this digital health where you will be learning about vendor neutral archives, ISI and so on. So these standards will be taught you in much more detail.
Yeah, so it is established and established that well in England.
Great. Thank you professorship.
bash, Professor jaiprakash or professor Ambedkar if any one of you can explain some part of, you know, case studies and capstone projects, which are covered in this program, get a lot of queries regarding these case studies and capstone projects.
So, let's see some of the projects, remember,
so, for example, the one knew the medical way segmentation, there are some problems related to ECG data.
Like how do we classify this data, right?
So, mostly related to medical imaging, the data that is related to the medical domain, where like you are given the existing like
the that already available publicly available data, and what we, what we'll do the project will be on various methods right, like, we share the code that existing code and you play with the code, you try to understand the code and you try to
run the code and change the code.
So, that you know, you can map all the knowledge to this experience, right, like to the really nice product. So, President progress, you have any other product, so, to remember, yeah, so, there are a few projects where you will be taught, where you will be doing
simple projects like COVID, segmentation, automatic segmentation of COVID data and so on. And they also have projects on
Yeah, I think your projects on ECG or eg signal analysis, and also how to how to do deep learning based based algorithms for speech processing, and I am right. And also you, you're going to have projects on imaging and vision aspects of
meaning and vision aspects, where you will be you will be you will be asked to ask us deeply given your networks to
do tasks like segmentation, segmentation and so on. But basically what, so, in this context, what what is important is, you will be given the data completely for open source data, and that open source data, you can download that data and you can also download the code, okay. And now, the whole, whole,
the whole expectation is that you will run this code, you understand how to how this code works, and then
then you'll also be having one on one interaction with the activities, says that,
if you have any questions on the different aspects of data or code, it can be discussed in much more detail.
Thank you to Professor jaiprakash. With that, we come to all the questions. There's one last question from Dr. Manu Shetty. Regarding can we collaborate with IFC for new projects using new data generated in their respective institutions?
So I can answer this. So, Doctor, man, the, there is some problem here, because we don't want to get into these IP issues. So, it's always better to have to work with open data, okay, where you where there is no IP issues that's coming coming between the two Institute's
right wait. And so once you've dealt with the code, and once you run the code, right, you can adapt that to your data.
I hope that answers your question.
So with that, we cover almost all the questions. So rubbish over to you, if you want to add, I would request all the participants there's a poll which we have launched,
to participate for the poll so that we get an idea about how you take the session, and what are your interest levels for this program, so that we can get in touch with you so I request all the participants to participate in this phone over to your Amish.
Thank you so much. Thank you, Professor Becker and Professor Jeff burger. Thank you so much for this session.
Definitely we will reach out to the participant whom we are not able to answer something, you can have our contact details, which is mentioned that you can reach out to us and we'll see how we can take it forward from there. Right. But thank you so much for attending the session. Thank you, professor.
Thank you, everybody.
To save this
so in this session thing we have to close
are you there?
Yeah. remise the session today? Sure. So, yeah. Okay. Thank you, everyone. Thank you. Thank you.
Watch the entire interview here https://www.youtube.com/watch?v=oNwiwdQGGtY