00 : 00 / 05: 00 (Preview)

This discussion has ended. Watch the recording on Medflix app,

AI in Dentistry

Dec 16 | 1:30 PM

What once seemed like science fiction is now a reality in health care. Specifically, in dentistry, AI is helping dentists identify, predict, and diagnose dental complications with higher accuracy. Additionally, dental laboratories and educational institutions use AI to augment their healthcare delivery experience - making AI a force to reckon with at all points in the healthcare value chain. Let's understand more about its incredible applications from someone who believes in applying digital solutions to dental science. Join Dr. Nivedita Tiwari as she covers the applications of AI in dentistry, live and exclusive on Medflix!

[Music] i welcome you all onto behalf of team netflix today we have with us a dentist who believes medicine and technology must go together to give exceptional services she's the clinical director at logie ai world's first quick oral screening module and is now used by india's major dental chains and prominent oral care brands a passionate writer of articles she has contributed to a number of oral health journals including aerospace dentistry stress and odoral health periodontal medicine and digital smile design doctor nivedita tiwari i welcome you ma'am hello so hi yeah yes so i will quickly run a poll before starting the session all right i know everyone is excited for today's session over to you dr doctor and thanks thank you samantha for the great introduction and thank you for inviting me for your show so uh hello everyone uh i'm dr nivedita clinical headquarter and uh i specialize in building clinical ai projects you can say and how i started my career is by pioneering two ai projects that were related to dentistry so that's why i say that i have been spearheading that ai revolution in dentistry but i also work uh for clinical ai projects uh with government of telangana uh related to general medicine uh so you know when people think that i talk about ai in dentistry they think that i'm a coder but no i'm i'm not a coder i don't have a background of a coder but uh like i'm a doctor who uh who just was just using our medical knowledge to build this innovative technologies that can be really impactful so saying that today's webinar will be about ai in dentistry i'll talk about my two projects uh about how i have built them how i've applied ai and also i'll be discussing about you know career opportunities for doctors uh out there how they can you know take advantage of this uh new uh you know ocean of opportunities opening up so saying that i'll just swipe to my next slide now you know when people like a lot of like i am sure a lot of our audience are quite you know knowledgeable of ei but i'll just give you a shot over we will not take too much of time on this that we of course know upskilling is required in any medical professional but we take up skilling only in terms of clinical knowledge and not much on the technology that is evolving and coming up especially in dentistry we work so close with all this innovative technologies but we don't spend that much of time into understanding how these technologies are being built so ai is the skill of the century you have to know that if you don't then i i feel sad for you because you have to know this is the skill of century automation is the future and you should be knowing something about ai uh being futuristic of course it is trending uh digital practices you know post covet everything is going digital we are talking about telemedicine we are talking about digital therapeutics everything is digital but that is where you have to be smart and you have to be technically sound interdisciplinary career now uh there was a time when most of the doctors and engineers they were all just you know very streamlined to their career they didn't wanted to uh collaborate or work in a way that you build a team in a way that they can you know work together and create something but now that time has come that even a doctor has to sit with the engineer an engineer has to sit with an mba and be able to build something so that time has come up what are the career opportunities so i will tell you very uh directly that for doctors there has been a speak in demand especially in the ai sector because multiple big players like you know amazon google microsoft they are building ai projects related to healthcare and they want a people with medical background but the problem is that since the people with medical background don't have enough knowledge on ai uh they are not being able to you know even know that there are such openings out there so we see we don't see that a lot of people are aware of these opportunities because they don't really understand what is ai and how ai is really working so one of the you know very big roles that is coming up for medical professional in this big shot companies are like clinical ai consultants data analysts ai product managers ai project managers medical data annotators what is data annotators i will come to that eventually medical data scientists so these are key rules that are that have started coming up and you can search these key roles on linkedin they will show you the number of openings that have come up uh yeah so what is ai so ai like uh is just replicating human senses you know how we think how we see so we just this is a technology that is just trying to you know replicate the human senses to put it in a very common layman's term so how is this evolving is before uh in computer science we only used very traditional algorithms meaning we we used to tell them do this and then that algorithm did but now the algorithms have moved to such a level that you can train them and then they can you know they have their network of understanding by of which they can actually predict diagnose and things like that so there are multiple levels of intelligence like narrow intelligence general super intelligence currently we are at narrow and general intelligence level on this technology we are not at super intelligence so narrow in general will probably mean to the level of human understanding uh super would mean it's going beyond a perception of human beings so that is still little far-fetched but people ambitious people are still working for it so what is ai and what is clinical ai so you know what is a ai is a technology what is clinical ai so what we do is we use ai into potential medical cases to ideally prevent disease detect any changes in the patient medical conditions uh diagnosed patients and whatever personalized treatments whatever so whatever related to in the medical sector we call it clinical ai so applications uh what is there so i will give you a quick uh you know quick overview of everything uh not directly jump into uh dentistry as and as in such because they are all kind of interrelated they have both major you know if you look at it so there is either product innovation where you know you are building a product uh it can be any kind of a software or a hardware that you are building it can either be a process innovation so in medicine like if you see we have multiple processes to be followed like digitizing doctors handwriting key prescriptions so they come under process innovation so how do you build such projects how do you know that how what is the life cycle of these clinical projects that we do so firstly and i will also tell you at what steps uh role of a doctor comes so to put it out very clear that no clinical project can be successful without a doctor they need a doctor if the technical background people think that they can build something without the help of a medical professional they are very wrong and if a doctor think they can build it without the help of a technical guy they are also very wrong so we have to understand that these projects require three stakeholders they need one medical professional uh they need a tech guy of ai background they will need a data scientist uh but the data work can also be done by the medical professional provided that he's little trained on that sphere which is not very difficult uh lastly uh then there is this last step that comes where we need somebody from a marketing background so that is something else i'll come later why we need that so firstly step one is you identify the problem so not all problem needs ai to be solved there are few rules why and where you have to apply ai you cannot apply ai everywhere you have to collect good data so data can be in terms of for example if you are like you can see in the image this is one of my oral hygiene projects where we created a caries model uh so we collected data images with kds in it so similarly you can collect data of x-ray city scans uh you know prescription modules whatever whatever data clinical data you have then you have to annotate the data annotate means if you can see this bounding boxes green boxes that i have drawn is just like drawing a bounding boxes across those areas of interest a doctor is trying to give a diagnosis for after that what what the ai does is we take this data we train the model we create an algorithm so once we train the data on annotated data sets the model learns that uh what is exactly a carries and what is a what is not a cadence what is a normal what is abnormal what is suspicious what is not suspicious so it kind of differentiates just like a human mind for example when we read our medical books and we go through the images we know what is normal what is abnormal similarly just at that level a i will also learn then the product is deployed now this is the final step which does not come in the scope of today's uh discussion but i am giving you a quick overview so that it kind of completes the journey so once you you know build that product you have to go for clinical trials for the clinical trials you have to you know then test that product on field and check its performance its accuracy and then there are multiple matrices that we check to understand how the product is performing is it performing uh as good as any human or is it below a human's performance is it acceptable because being a healthcare we cannot take a chance of giving a wrong diagnosis or even a wrong screening to anybody since it carries a lot of you know danger then you have to get it validated by number of regulatory bodies they again vary with respect to countries like india has a different set of regularity bodies europe will have america will have and then you have to sell your product so when you make such products especially if you have your own company like i have mine uh we you have to understand where this product can be useful where can you sell it so that is another level so ai comes at four domains i have personally worked on first domain which is computer vision computer vision means trying to replicate the eyes of human whatever a human eye can see and perceive computer vision is trying to achieve that natural language processing uh classic example would be you know a predictive text that's come when you're trying to put an email to people you will see some predictive text starts coming that is a natural language processing ai use case uh speech recognition classic example would be a voice command system like alexa siri they work on that level and robotics is basically if you guys have heard about these drones that have come up to you know deliver blood samples and medicines and other kind of surgeries robotic surgeries that have come up they come under that division i have personally worked at one division which is computer vision so seeing the scope of today's webinar it's not possible to cover all the domains because it's huge it's vast applications are innumerable applications are mind-boggling so keeping the scope of today's webinar i will be only discussing the computer vision and the projects that i have worked on so when i said that uh step one is to identify the use case when do you use a uh when you use ai so first criteria should be that when human can perform the task in few seconds if the human can perform the task in few seconds so that you can train the model in that way and then it can also perform when it is impossible to write down the rules so for example when you say somebody has a oral cancer you have to take multiple factors into consideration what is his family history or even if to say that somebody's diabetics or okay let's not go to diabetes that is not related to exactly computer vision i can tell you uh let's say to com cancer oral cancer so for example you say that if somebody has a medical history of personal history of tobacco chewing whatever their family history so there are multiple factors we consider before even saying uh just by looking at the patient mouth i'm not talking about diagnostics just by looking uh just by seeing and even thinking in that direction so similarly for such use cases we should use ai and when it is possible to collect sample data so if i want to work on oral cancer or similar projects lung cancer i should have that enough data with me so how such projects happen first you have to you know collect the data with proper standardization this is where the role of a doctor comes they have to know what is a good data what is a bad data can they see the pathologies in the image can they understand is the image good so this is called as data cleaning then you have to do data labeling and annotation basically you have to sit on the data and you have to give your diagnosis on the findings whatever you find then you start creating initial ai models so how ai works is just similar like a human mind just exam for example when we are in our first year we don't know what is medicine we hardly have any good concept but as as we move to the fourth year our brains plastic plasticity starts working and we can you know start connecting everything and get better day by day because every day of the year we are learning something and that's how eventually we become a professional we grow and then we become doctors so similarly the ai will work like that it will first create initial ai models then you keep training with good data you keep level so that process will continue continue continue once that model uh has some kind of you know mattresses like uh it is it receives a kind of a level which is acceptable then you start doing lab-based initial trials then again you retrain the model to get a better accuracy you again perform the field trials and then you deploy your product and you publish your results so if you can see the time it takes is very long so ideally any ai project takes 12 to 24 months of hard work you can say because we are trying to train it to be as good as any human uh so what are the so this was this uh thing how do you measure that the ai is giving a correct data so there are certain formulas we use like true positive uh true negative then actual results so true positive basically means uh for example if you see an image and you see that a doctor predicts by seeing the image that this person has a cancer and ai also predicts this is a cancer that is considered as a true positive what is a false positive uh for example i say being a doctor that no this is not a cancer but ai says this is a cancer this is a false positive it has given so similarly we put down the and false negative if a doctor says that this is a cancer and ai predicts this is not a cancer so i know this is kind of confusing but this is how it is this is how we formulate and it this is called as f1 spores basically so f1 via f1 score we try to understand uh how to you know measure the accuracy so once we collect the data we divide it into three sets called as training data sets validation data sets and test data sets so in the training we keep it at 70 percent so for example we have 100 images we'll keep 70 for training 10 for validation so for 10 we will be training the model for 70 of the data once the model is trained we will validate it that how good it is performing on ten percent of the data and then on twenty percent of the data we will test it just like uh a child like just like us going for an examination we go we study whole year and then we give an examination item there that is what test data does so how how do we so you know i get this question that is uh ai will replace doctors and i tell them that this is the most baseless query i uh that people are having especially medical professionals because ai will not replace doctors medicine will always be with doctors but yes i can say that doctors using ai will replace doctors not using ai to be honest uh because this technology is so awesome it is so good and it is just going to open so many opportunities for doctors to you know uh especially if we have this problems of unemployment or problems of over working where people even are overworking or are they unemployed we are getting a lot of opportunity in this sector where we can help and also we can give service so there are three kinds of systems in ai first is human out of the loop i'm sorry i can't see in the image i'll just check on my laptop so human out of the loop system where ai is just working independently for example of their numbers number of things over there you know like election prediction and other uh it's basically not medical use cases i can tell you human in the loop system again i can say cancer prediction because the kind of risk cancer carries and so there always be a human to cross verify whatever ai has found human on the loop system can be facial recognition systems like uh if you using an ai to uh you know understand who has attended today's uh work or meeting so that also needs to be you know monitored uh simultaneously with a person with a human setting so uh some most common clinical use cases are screening or diagnosis so mostly it's on image based or sound based decision support systems like prescribing medicine or giving a support to somebody's diagnosis uh and also very uh this has come up really uh right now is predictive insights that is mapping of disease and progression occurrence so i will now quickly take you over to some of my projects that i worked on this is the kds project that we tried to understand how we can detect caries with help of just smartphone based clicked images this is not my use case but this is a very common use case that has started coming up and very big american companies are working on this which is uh detections on opgs uh they basically detect all the pathologies on the opg automatically this is also one of my projects that i'm working is oral cancer screening where i kind of detect and understand what is the level of uh suspicion suspicious lessons that can we can identify in a person's mouth at the earliest and so the idea was that we give this tool to uh you know remote working individuals who are at remote areas and we know that we come from a country where eating tobacco is very normalized they considered it to be very normal they consider it to be a lifestyle pattern that they follow a behavior so seeing that and understanding that not not every doctor can visit a remote area to screen people's mouth that's how we started working on this project that if if a remote worker or technician a field worker can just carry a phone uh click the image raise an alarm of suspicious legends and then eventually take a note of all these cases and and refer it to the doctor virtually or refer it to the nearest government who can provide them care uh also we have worked on multiple smartphone based clinical ai projects so one is oral hygiene screening there is another cataract screening that we are currently doing with government of tamil nadu and we uh the same use case where you just go take a photo identify the person as a cataract or not and then refer them for surgeries uh then hair analysis like a smartphone but here we also take a smartphone and a lens and then we try to analyze the density of the hair so yeah thank you that was my uh presentation my work and i hope it might have given you some uh knowledge or some you know insights of what i'm doing and i'm open for questions now thank you thank you dr nivedita it was really interesting and i'm sure there are many people who are curious and are curious to know about a lot of things it was just a just of overview of few things i know there's a lot to learn a lot together yeah um okay i'll just read out few questions um okay so actually i wanted to know i was curious um like diagnosis in ai so you know how important it is in our practice diagnosis because it the treatment plan will all depend on our diagnosis so diagnosis has to be perfect has to be clear in our clinicians so how will ai help us in future sure so you know one of the things i see with people is we should not be very over ambitious especially with technologies and we uh always think that like i said in one of my slides that in medicine ai will always work with humans so it will always be like a decision support that means if ai is diagnosing something it it is getting confirmed by a human also human in the loop second we always do it for screening screening and identifying at the earliest so we usually never we never want to uh never i i don't even see that as a future to be honest where uh you can leave the whole of the diagnosis on the hands of ai and leave it because that that will also never get it will always be an assistive to humans also like um can we use this and pedo patients pediatric patients maybe cooperative pediatric patients yeah so uh i'll tell you i'll tell you one very good use case see we have doctors in their clinics and the doctor is waiting for a patient to come do you know that maximum moms even they are from a good family and they are earning really well they don't know what is really uh you know uh this early childhood careers they do not know what it is they don't know that i have to take my kid to a doctor the moment i see something so there was this one use case and there are multiple such use cases so the idea of using ai is for screening that's the use case i have worked on where somebody can use this uh you know screen because people you know we always do something that is we see on the internet we see on the but we will not take because this mentality is there in most of the indians you know we are not going to a doctor unless we feel pain unless we feel that okay now we are about to die we will not go so the idea was that how do you create awareness so i can say ai can help us in creating awareness telling people no give them the first hand of information let's see you have this problem uh you know generate the alarm please connect to the doctor go to the doctor because in most of the cases these people they don't even know they have a problem that so that is the perfect use case i feel wow that's right so we have few questions i'll just read out to you um okay uh dr yasid is asking um can you please enumerate difference between machine learning and deep learning okay so uh basically machine learning is you can say it's the you can say it's an umbrella okay and whatever you keep under the shadow of the umbrella is deep learning so it is just uh you know a sub classification where there are number of numerous types of neural networks involved so it's a basic more advanced form of machine learning that's all i can say okay um okay dr haroshini is curious to know like how to get into this field after bds oh yeah so that is a very common uh question i always get so firstly uh may i know your name sorry i didn't get to dr harshimi dr harshiri okay that is a very uh common question i get where people ask me that do you have a course is there a course i tell them there is no good for such things so what do you do if there is no course available firstly uh it depends upon what are you trying to achieve is it that you want to open your company or you want to work in corporate uh big corporate houses like uh amazon and all so what you can do there are two things if you want to open your own company uh you should start working with startups who are at an early stage and they are developing in ai some medical product if you start working with them you will understand how ai is working so you know i'll tell because if you see in even in our medical education unless we don't really open our clinic or even after engineering which until we don't go to a co go to a corporate uh office we really don't understand how things work so just by doing a course even for people with a technical background even they don't understand how to build such ai projects to be honest that is the truth why because they are not until unless you don't travel that whole path yourself until unless you don't open your own clinic and start seeing your patients by yourself you don't get that learning that understanding so for that if you want to open a company what you can do is you can start working with you can definitely do some theoretical courses that is available online to understand everything about like this everything about ai which is just theory based but to understand how really these projects work what are the challenges you will face for that you have to be associated with similar startups and to see even if you want to apply for corporate uh jobs in similar place i think this would be the best way to go for it that you start associating with companies or even with your friends or who have this technical background and start working with them start uh doing courses with them because we currently don't have any framework as such it is coming up the governments are sensitized now and they will bring it but for you to be on this you know um i can say to be above the race this is something you can start with also linkedin is a really great platform so apart from doing some online course you should just follow linkedin with you know multiple uh such great connections you will find who are working on this project and they will add to your learning so you can do that great dr harshini i hope this answers your question um okay we have uh this question um how to bring technology into private practice how to bring ai into the private practice you buy their products and you feed it i was like we are taught to like you know practice practice but there's every day something new happening new technology how to be updated it's how it gets really difficult but i think it's fine but i just i even i'm curious to know how yeah yeah so i'll tell you two things first uh you have to be very open-minded uh you have to start following uh these pages and like i said linkedin is a really great uh place to follow these cases now actually you know the problem is not how to the problem is that we mentally have been trained not to think beyond our box not to think uh above that you know that wall that we have created our head so for example i'll tell you when i was in my internship uh i was the only girl in my internship here who started working on digital smile designing dsd by coachmen if you know so i actually conducted a research study to understand what is awareness of a digital smile designing among uh doctors of dhavangiri dentists you'll be shocked not even 70 percent of them knew what is dsd uh when i did that research i got it published in one of the state journals of karnataka that's you know and when i was doing that study most of the people were like you're wasting your time you should actually just prepare for mbs and i was like i want to explore options i want to see what what's there out in the world so i would say that thinking is the only barrier there is enough of knowledge out there there is enough of a dental companies you should get in touch with these founders of this you know uh like densify sirona and these other companies are there such big tried connecting with them see what is out there it's you are in world of internet and knowledge is free so there is no excuse at all i would say that i hope that yeah that's so true like even i remember my days like when professors used to teach us like you should always have a thing like you know okay this yes i have a vessel but you should always ask why why is it like this why is it like like yeah so yeah oh okay we have a question from dr madasir uh how can i start a project of diagnosis uh so for diagnosis uh so you have to first understand what kind of diagnosis you are looking for maybe uh what area of medicine you are thinking about that is something you have to understand first so it depends on that because uh there are so many diseases what kind of diagnosis you want to act on if you want to act on multiple diseases i think it will be little difficult to begin with ideally you should begin with one kind of disease and then you should move on to different diseases that's right we have a question from dr aminal uh how do we get in touch with you for further details follow me on linkedin i think that's my i can share my linkedin link and my email id you can anytime contact with me i'll be happy to help i'm always happy to help all the medical professionals uh and if they reach out to me i always i always uh i'm always there to help you not a problem we have uh two people on stage dr yasir i am just accepting your request please uh turn on your audio and video doctor yessir [Music] yes hello uh good evening ma'am good evening hi this this is dr yasir actually i want to know the difference difference between this artificial neural network and convolutional neural network it's just just so confusing at some times because they are overlapping yeah so you know and cnn i would again say uh they're actually kind of at the very technology side of it i would say they are the same they do the same thing which is you know mimicking how the brain is working but to go at a very technical side of it is something uh actually out of my scope currently but i can tell you what they exactly do is so you know how this medical projects work is when you start training the data it goes into a box okay it is called as a black box it consists of various neural networks with which has nodes just like you know human synapses where we have a node we have a neuron and amazingly that structure is built in that way just the way a human neurons transmit messages so they work on that similar scope so they are actually basically same uh and they actually basically do the same thing but i will i will not be able to tell you in details what exactly uh has because that is not in my scope coming from coming from a clinical background thank you thank you um [Music] we have a question actually a comment uh dr chirag uh i think ai would be a breakthrough in our country as most of the crowd follow um ongoing years like the old tradition i thank you ma'am for showcasing some light and exploring stuff and breaking the stereotypes um there was one more question like what is the scope after orthodontics okay so the best scope i would say you know uh there has been multiple uh startups if you know them about like there is tootsie and all they have started coming up with this innovative solution where they are actually you know so of course the first option is of course opening your clinic that is what most of the dentists are we are doing in today's time but i also see that there are multiple startups have come up where they are you know giving home care home delivery so in medicine one of the things that's coming up is home delivery uh care care delivery at home so uh for that a number of you know virtual digital solutions are coming up and they are requiring uh especially autonomous uh they need orthodontist for such so you should start applying over there this is one of i think this would be one of the great app like you know newest applications uh that you can work with such companies apart from your practices so that's something i can suggest that's right we have dr nidhi um she wants to know about the future so i'm sorry i think i got the question wrong is it did he mean a orthodontics in ai or in general career i'm sorry after orthodontics like what's the future of it yeah okay not related to ai right uh nothing because that aiu sector is different that is a technology site so any scope of see again like i told you there is definitely a big scope because first of all uh orthodontics is now evolving to be delivered at home so we need smart solutions out there people need more smart solutions to screen the crowding screen the arc length and whatever parameters are there to even monitor so you know most of our solutions that is there in orthodontics it's completely clinic based so you know to bring out the solutions that can be given at home also that is also something a big application out there the process is again same please reach out to the companies who are working in space of orthodontist i can recommend you many uh you can reach out to me i can recommend you there are multiple uh international companies who are working on this space of orthodontist you can reach out to them start working with them understand right um okay we have a question from dr nidhi like um she wants to know about the future of ai in endodontics endodontics okay so of course there is a future but uh again the applications are really vast in endodontics uh currently like i shared with you that whatever is happening currently what we were doing is uh basically general related to general creation of awareness but if you are very passionate about solving a problem if you are very passionate about thinking that if i was not there and there was this machine who could do this for me you know this machine who could help me because now with my explanation and webinar you can understand where ai can go to what level it can go if you think that there is a machine out there which can actually help to solve your problem your daily uh problems please go ahead go ahead and collaborate with people start working networking that's right so since there are a lot of dentists here um i just wanted to share a few things that we are going to start a dental club so we are open to faculty suggestions topic suggestions uh you can comment in the comment box so yeah this is a very interesting question from dr chetna is there any role of ai in dendal laboratory yes multiple multiple i think there is a one very good pioneering company called as pearl which is working on uh you know impression taking on how to take an impression how to recommend whatever uh the size of the crown placement and everything building so that is one of the pioneering companies working in that area which is going to automate the whole of the process of how to you know take the impression then then you know automatically design a crown a metallic ground and whatever and then they will eventually you know get it transferred to the patient with and they claim to have reached a really good precision and they are working even better than uh technology so i would say dental technology they're amples and amples of applications that we can explore we can start working especially in india and you know most of the time the problem with indian market i would say is we have a lot of questions we have a lot of use cases we have a lot to you know talk about and explore but nothing much is happening here because because of multiple levels at multiple uh you know levels of deployment and adoption of technology so especially when i when people ask me all this question i say yeah this is happening but it's not happening in india it's happening somewhere else so that's how it gets a little you know uh it's not reach it's not reached out to us yet that is one of the problems but definitely multiple uh problems are there that can be solved actually a lot of people have this question okay i've completed my bds so how can i join our ai company or how to go about a click how's the whole process i know like i always get that question so even people were like uh like should i start but my you know my work commitments are so huge that i'm not able to start that course i i would love to start this course as soon as possible uh so you know like i said i have already discussed with you all what are the roles there for medical professionals you don't have to learn coding people think you have to learn everything about coding and there are no building part is responsibility of engineers you leave it to them you will have your own role you will have your own applications where you have to use your medical knowledge you just have to understand uh how this is working how am i training uh this just it's just like you know training your own child so the course again i would say at since there uh since there are few courses out there which actually just tells you theoretical understanding of ai so you should start attending and get to the basics of it uh start involving with companies or startups who are at an early stage so like a lot of people try to come and try to work with me but i tell them that we are not in early stage so an early stage startup is basically at an ideation stage they will like to you know work with you to develop something but if somebody has already developed they are at a different stage there are advanced stage they are looking to deploy the solution for clinical trials and market validation and other stages so when you should ideally start a lot of searching working on this you should try approaching and working under people who are into uh who are into this so you know the concept of internship let's not just keep it to medical and dental colleges let's keep it online uh network on linkedin try reaching out to people and start doing internships with them you will definitely learn that's true oh i had a question actually uh you spoke about sound based diagnosis so i was just curious to know how how it works so sound based diagnosis is something that has come up recently where still it is at a research level you can say it is not deployed yet so there was multiple attempts even for covid to understand that if somebody is coughing and just by listening to the cuff can i say that this person has covered or can i say if a person is having a productive cuff or a non-productive so these were some research that's going to happen and it is still happening it is at a research level i would say not really uh not at a level that it is deployed and being used in hospitals uh and in sound there are multiple kind of sounds that are coming up like you know heart palpitation understanding uh reading the sound of heart palpitation via technologies and reading the ecg these technologies have also come up and multiple uh you know uh this uh non-invasive uh contactless devices contactless like you touch them and then you understand what is your uh you know this ecg level and your uh whatever pulse and everything that comes so they are also using this technology but just by listening to the sound of a voice and be able to say that what kind of cuff it is or cover it is that under research it is still people are still working on this current but i'm sure it will be really helpful for the clinic so if you are interested you should find those companies you can reach out to me i can uh tell you about these companies and go and have an internship with them sure people were really curious to know like how to go about it and i'm sure you've just given the overview and there's a lot to learn and very interesting um so you know i'll tell you ai is a big blank board it's a big blackboard people are writing the formulas their people are writing how it's going to go eventually it will come out as a course to everybody it's not at now because every because the kind of courses we do to these days be it medical education engineering and everything it's it's losing its uh essence to you know there is one part of learning and there is one part of application where you apply something and you see it in real time how that product is working how that project is working so we are eventually moving to uh you know that sphere of the world and for that you have to be very alert you have to be always updated with whatever is happening read a lot read a lot about ei you cannot know everything in one day because ai is such a you know it is it is like exploring us it it is exploring humans basically how our senses is working every day you will see one research being done where somebody is claiming that we uh we found out that we found out this and then you want to understand okay how did they build it what you creating that algorithm it requires multiple experiments i'll tell you it's not easy it requires like i said that the neural networks that is inside the computer's mind it is just as same as a synapses of a human brain which is you know taking the information transforming its and we don't know which formula is going to work to give a diagnosis for example if there is a cancer in the patient you will see that to develop an algorithm to say that okay this image is giving a correct diagnosis ai is given it takes multiple levels of iterations to come to that level so it is a very evolving very exploring subject uh please have a look at it please be open-minded read about ai uh read about i regularly post about first up on linkedin if you just follow my post you will get to know a lot about it you can follow my connections to they are also very much similar uh thinking like me and you can definitely learn a lot that's wonderful

BEING ATTENDED BY

Dr. Murtuza Zozwala & 1072 others

SPEAKERS

dr. Nivedita Tiwari

Dr. Nivedita Tiwari

Dentist | Spreading AI revolution in dentistry

+ Details
dr. Nivedita Tiwari

Dr. Nivedita Tiwari

Dentist | Spreading AI revolution in dentistr...

+ Details

About Medflix

Medflix is a new platform by PlexusMD, India's most active and trusted doctor community. On Medflix, you can discover live surgeries, discussions, conferences and courses from some of the top doctors and institutions across the world. Join clubs in your areas of interest and access hundreds of amazing live discussions everyday.