thank you yeah thank you very much yeah welcome so i would like to talk about something that is very important the resistance of antibiotics and how we can actually kind of mitigate the risk that we will have more and more of those resistance in 1945 alexander fleming received the nobel prize for the discovery of penicillin actually it all happened you know with the incident in his lab he had some petri dishes with bacteria he was working on went for a brief holiday and when he came back he found that on some of the pity dishes a mold had developed where this mold had developed the bacteria were actually dying that gave him the idea that this mold so this fungi would actually create something some toxins would actually interact with this bacteria and that it is maybe something that could to be developed in a new truck and together with his cool collaborators foreign chain he received a nobel prize in 1945 for this discovery but already then so when he had discovered that he was aware of that that there could be resistance because bacteria proliferating very fast if their drug is not used in the right in the right time in the right level so he was very concerned about that and it happened to begun and be true and now is a big problem which we have to tackle let's look at that what has been the time before and after the discovery of antibiotics and it cannot be overestimated how this has changed the world and how it's changed medical sciences if you look at that it's just mind-boggling you know the average life expectancy before the discovery of antibiotics was around 47 years after the discovery of penicillin and antibiotics it increased by a mind-boggling 30 years through life expectancy around 79 years today unfortunately we see already resistance coming up so our you know weapon sharp weapon of antibiotics becomes blunt and it's estimated that by 2050 people may be more likely to die of infectious diseases because the antibiotics would not work then of cancer and that is called the medical climate crisis it would also lead to a massive reduction in the gross product of the world democratic product by two to three point five percent if we don't be able to tackle this crisis and in the covet pandemic one thing we also have seen there was a lot of antibiotics being prescribed to patients indiscriminately precautionary but that reduces the window of opportunity further because it creates more resistance and the patients coming out of the covet might then otherwise suffer from antibiotic resistance and bugs that could be superbugs now what are we concerned about and why is it so important actually to treat the patient in the right way and not only give antibiotics in this community so we have to think about the patient himself it must be the right patient the medical history plays a role the surgical history and the antibiotic history of the patient is very important when you prescribe antibiotics then you need to also find the right diagnosis because you need to know what is the bug the bacteria that causes this illness what type of infection what type of hygiene and original we need to also think about the right dose antibiotics has to be dosed in the right way so you can overdose or you're under those so you need to figure that out and you need to include those information you have from the hospital from the world health organization or from ucast and when you give the antibiotic the duration is also important too many patients you know feel good again they stop the antibiotic treatment by themselves although the doctor has prescribed it and what happens the evolutionary pressure on the bacteria is residing and bacteria become resistant then we also have to think about what is the pathogen we are talking about what's the bacteria that is causing the illness we need to think about the the antibiotic itself the right spectrum uh there are broadband spectrums and uh very selective antibiotics so this is very important to think about it the right struck thinking about resistance and about the right drug that it could be you know for this particular infection also allergic factors are important to think about here and in the end the question is should it be administered orally or should it be administered to intravenous injection now why do and bacteria become resistant to antibiotics well there is a mechanism in bacteria they grow very fast they have the ability to exchange gene material even if they come close they can exchange the same material there are spot mutations and that all creates very rapidly resistance to antibiotics then when you look into the situation in hospitals or you know in places where you know antibiotic and where the bacteria can be found you can acquire it through the ventilator you can acquire it in the hospital so there's site-specific resistance which is very scary and you have heard about superbugs in hospital spreading and then we also have to think about the person itself because in bacteria is interacting with the microbiome of the person in a very complex way when we give an antibiotic we don't want to kill the good bacteria which are part of your microbiome but we want to target the bad ones which are causing the infectious diseases now what we have created is we have created a knowledge graph where we integrate the knowledge about symptoms diseases infectious diseases so that we can find out what is the right disease we are talking about this is a belief knowledge graph that maps the symptoms and those disease information we have the disease ontology on the other hand we have the information about the bacteria about resistance which is already known about bacteria in this categories aware watch and reserve in the antibiotics that are being reserved and here we can actually help the doctor in the prescription process to guide them to administering the right antibiotic how does it look now when we look into this we'll see that these two graphs which we have mapped to a neo4j data base graph database includes this enormous knowledge medical knowledge which is already there about symptoms diseases gene and symptoms into relationship resistance patholomics that is the belief knowledge graph and on the other side to administer the drug we have different ways how to guide to the clinical pathway so to help the doctor to prescribe the right antibiotic here this is the symptom and we will show that in more detail my colleague ashoka will guide you soon that it shows you a typical scenario for that the patients come in will complain about some symptoms having fever having night sweat this type of symptoms is automatically interpreted it's mapped to a controlled vocabulary this controlled vocabulary using standards like the international disease classification system it becomes machine interpretable then based on this diagnosing the right infectious disease now the antibiotic that would most effective will be suggested and this antibiotic then can be administered either orally or to intravenous injection what we need to think about is the right care at the right time at the right price the right point of care from anywhere avoiding antibiotic resistance because we give the wrong antibiotics and we give the wrong antibiotics during the wrong duration so this is the point where you know rust reiterating the 6d model of antibiotic stewardship right diagnosis right disease causing agents very important to figure that out and then clearly giving the right drug de-escalating that by giving the antibiotic which is more suitable for the illness in the right dose and the right donation hand over to my colleague ashok yes if you can see my screen i will now offer the demo as roland has just said is that we have right now the digital twin and the digital triplet digital twin is the reflection in the digital world of the real world whereas the digital triplet is an abstraction of the real world into once again in the digital world but the difference between digital twin and the digital triplet is that digital twin is the one-to-one mapping whereas digital triplet is an abstraction of the real world into the digital world and i will take you through that say for example now when and this is i'm going to show right now is the digital twin of a physician when a patient is having some kind of chief complaints then he goes to the doctor and in this case have taken fever which is one of the most common one and we can find out that there are 474 diseases which are associated with fever and these are some of the areas where there are a lot of you know misdiagnosis now i add few other symptoms or chief complaints along with fever we have night sweat and along with night sweat we have given coughing blood and then we have also given hemoptysis now here what we see is that when i'm giving symptoms it has come down to three diseases that means these three diseases have those symptoms or the conditions in common and on the right here you can see are those are human understandable english in natural language where i said fever night sweat coughing blood and hemoptysis now hemoptysis and coughing blood hemoptysis is understood by the doctors whereas for common people coughing blood is much more understandable but you can see on the right where this human understandable natural language through a artificial intelligence engine has been converted into machine understandable code and here you can see that it has given three diseases pneumonia heart cancer and tuberculosis now heart cancer many doctors will miss it and what they are likely to do they are likely to predict as tuberculosis and the treatment for that and we have the detail and you can go here heart cancer which says the symptoms are hemoptosis then fever and then night sweat so you get that in a moment in front of you the chances of error is less pneumonia is a infectious disease so now i will show you the digital twin of a microbiologist from the knowledge graph so what we are trying to do is that we are mapping the human mind and human brain into the knowledge graph and here when i click pneumonia roland talk about the right diagnosis so from the symptoms we did right diagnosis now the question is that we need to identify what is the right disease-causing agent that is what is the bacteria causing this particular disease as i said that 26.5 million antibiotic courses per year given to patients who are not having tuberculosis and the patient number is 5.3 million patients per per year and here you can see the disease causing bacteria are those and we can see on the next column what are the antibiotics resistant to this particular bacteria so that means if we use any of those back antibiotics for this bacteria then there will be resistant that it will not work or for example streptococcus caucus pneumonia these are the drugs which will not work and here you can see what is the digital triplet and digital triplet as i said here there are two abstraction one are one is the microbiomes which you can see microbes and then the diseases infectious diseases and here what is very interesting is that as i said that in this abstraction we can find out unknown knowledge let me repeat when we are talking about digital twin what we have is that we convert the real world into digital world and we can find out what are the hidden knowledge practically in human brain those are called tacit knowledge so we convert acid knowledge into explicit knowledge and store it into the neo4j database or the neo4j knowledge graph and here you can see the infectious disease which are in this space very close to pneumonia which is meningitis and practically as a matter of fact step to caucus pneumonia when it goes to the bloodstream then it is likely to affect the brain and cause medic uh meningitis so for a doctor this is for many doctors this is an unknown knowledge and then what we do is that and here are some of the here is i just talked about streptococcus pneumonia and these are the antibiotic i'm just showing you in the neo4j knowledge graph the cipher comment here which fetched all those things now while roland was talking he talked about intrinsic resistance intrinsic resistance happens due to the gene transfer and when it is a permanent resistance that cannot be changed because it is now genetic but there are another type of resistance which he talked about is site specific which are specific to hospitals or within a hospital a clinic or even it could be an icu so here from the icu data we have taken the antibiogram data which are sensitive to the culture data and here what happens is that in that what we do is that we take the culture data and from there we did bayesian probability network and from there we have organism and there are different antibiotics which may may not work which is not a intrinsic resistance but this is an acquired resistance in a particular setup like community acquired pneumonia or ventilator associated pneumonia okay there are quite a few number of people when they are in ventilator they acquire pneumonia from the hospital itself from the icu or from the ventilator so this is a probabilistic graph where we have that and here is an example of the result is that here i have given that for this antibiotic in the hospital icu which antibiotics are likely to be resistant and likely to be susceptible so here it says that celita is 100 susceptible so it is better and this is a data from pediatric icu in one of the tertiary hospitals in india where we have the okay and here is the antibiotic stewardship antibiotic stewardship is now this is about precision antibiotic therapeutics so what we do is that here there are few cases to make sure that we don't underuse or overuse antibiotic so here it is a covid case and we have in the knowledge graph the evidence and here the evidence says that particular antibiotic should not be used you can see it here in the interpretation our findings do not justify the routine use of azithromycin so what we do is that we provide the and then the stewardship committee they reject this no it should not be used then there is another case where we have a case where somebody has offered a broadband i mean a antibiotic which is of broad spectrum so here we are trying to change looking at the evidence we change the antibiotic to a antibiotic which is likely to cause little harm and that is why we are calling it precision antibiotic therapeutics and then the doctor he approves it okay so the point is that when it comes to the stewardship committee the stewardship committee they go through each and every antibiotic prescribed and then see whether we can reduce the broad spectrum to a narrow spectrum or can we change it from intravenous or intramuscular to oral because all those things will reduce the chances and possibilities of this particular bacteria becoming resistant to some drugs okay so here it is after changing the drug the recommendation it is getting approved the last one is a two-year-old patient a child who is having pneumonia and here it says that for this this particular a broadband is sufficient and then we approve so this is approved so what we try to demonstrate here is that we have shown the digital twin and the digital triplet of a doctor's mind where we do the right diagnosis and then we went to the digital twin and the digital triplet of a microbiologist's mind to show that for that infectious disease what is the disease causing agent and for that disease causing agent what are the right bacterias to be used and if you see here in this we are also talking about ivdd ddd is defined daily dose of that particular medicine which is recommended by who and whether it has any oral substitute if yes in this case for the first one there is no overall substitute and what should be the day of the therapy so we offer all those information to the physician or the practicing doctor at the point of care so that the consumers the chances of error is less and the antibiotic resistance is controlled thank you
2021-08-04