Understanding Health Informatics – Ashraf Farrag (#3 of 7)



thanks everybody for us here that this dialogue job providing his spoke student someone successful so I'm going to address a kind of my prospective as not a faculty person with somebody that interfaces with faculty here and how what what's involved in health informatics agency so I have a research analyst that's my product title that's he track sensitive musical building over my house for me in particular I have ties to the main campus through I TS research computing my old apartment technically I'm on loan to nctracks Institute of your research and interesting part is everything that I worked with in terms of data is in our hospital system so the three unique asylum is a funeral that I have to jump between occasion even parachutes but usually you guys just as at one so I'm a product of this lovely shade of blue no offense the other one ends well enjoy collaborating with them so the one point I would made on just my background two words there look at one is this bioinformatics which is dr. Andrews apartment in health informatics there are two different subjects this time set the stage make sure everyone understands that but what we do in the health informatics program is not necessarily the same as bioinformatics witches street somewhere but they are very much related and actually in the future that are going to be very much time together bioinformatics Neil's long lines of cell biology genetic analysis those types of information topics working gene sequences things like that in health informatics we kind of zoom up or not looking at the cell looking at her patient obviously what happens in a cellular level is important to the patient but we're looking at possibly the whole picture and as the future continues to join the two of them I imagine that is helping will have to absorb some of those things to be but primarily what we work with in health informatics are things that are related into the nation so imagine you go to the doctor nowadays you have some kind of electronic health record with a record your blood pressure of your pulse respirations all that kind of stuff and then you have your visit with the doctor and at the end of the visit will type up your mind so that what about you how your lovely patient they love saying every time and this is typically a piece of text it's a letter no so those are the kinds of things that we work with here influence under the health of heretics discipline is this new thing this combination patient recorded data everyone's got a cell phone we've got a Fitbit we've got scalar so this is kind of a new thing in health informatics is coming down the road where physicians are going to start getting at data from the patient and add to professional devices of calibrating professional quality to have came out this more constant stream of information communication so those are kind if you can kind of think about the jewel in the eighth with two fields asking you to split them right now but again I'm under then they do tie together and lo and behold where I work we are really the intersection of that right now our motto is from the bench to the bedside so tracks again when peace TV programs multidisciplinary it's all over the campus we strive to take research has done at the lab bench doing the gen sequencing drug discovery and bring it all the way to the patient's bed the effect treatment is quickly impossible so let's kind of work to get at this notion that why are they renamed differently the future is of the two will join I hate slides any X that slides off you know practice up pretty quickly and kind of hope that little bit of dialogue so what do i do i've talked about all these institute Simon the corn what I do is access data that is contained owned by the health care system the hospital in that business large database called kill anyone else health essentially we have all of the health records organizations and unc hospitals and ceramic clinics sorry 2004 in Qarth day and that's essentially anything has been recorded in any sort of patient nurse patient doctor revision provider period interaction so we have access to huge information repository they're kind of two things that we have in there on this structured data imagine rolling back for traditional database numbers lab values things like that and other pieces unstructured which is what the clinician or the nurse such a social worker put it into a note which isn't excellent like a book and I don't want I think that we will talk about Donna structured data to to what she does but we tend to focus on structured data where we just extract information researcher comes to us and says hey I'm going to spend our movies and patients can you find us patients can you find us information about the Persians so we go in do our data mining provide them an extract and along they go through the research get the grant funding etc etc etc the other resource I am probably the most is unstructured data we do have some facilities where we do look at those projects in particular I think kind of the bridge to how healthy venison inches soft on your ass that's probably one of the areas i think that would you significant overlap for instance you know look at the patient disease remission want information about it background that's something that when you're designing in your tooling to extract this information out of those notes knowing where you know I need cancer ontology list of medication terms that's something potentially where librarians would have a good handle the various resources for that that for these projects could be of help so by a large kind of my perspective the distillate is I think that that would be one opportunity for help pubmed i think is going to be a significant resource in terms of doing these types of projects with unstructured data marrying kind of the data we have locally about the patients to what are we seeing clinical trials best practices for searching disease specific research so that's kind of my taking in terms of what I do have potentially how people in hsl can interact with it so largely what I do support research also supported us for quality improvement not so much hospital operations so all the things that Jonathan mentioned roll up into this what's actually happening right there on the battlefield so to speak informatics the other thing I would say is I have kind of as a chip graduate I tried to foster student projects his early student once and was in need of a project so over the years I have tried to bring students to tracks and work with them if they tend to work now second fourth room that we have lovely room through the brainstorming there there in this quiet location so that's kind of our home that we try and cook up these projects some we've been working with clinical data so we can work in they have challenges where they have a son of data and they ask a student to try and do this task to disinformation extraction plans so those are kind of the things where I think that those projects will be helpful for them to be able to quickly find resources like we're doing something with cancer okay what are all the different cancers resources databases etc that we could mine and then mash up with the existing in so that's all I have in terms of presenting to you of the big questions or clarified has the cops gone down to hold of the information you know one time to remember some scientists talk about you know they the cost of downloading the information and then being able to sort it because they're so massive I assume that you know that that's improved in this might have been five years ago at the University Florida media kind of operated tracks on a fixed rate we don't go by volume we go by the complexity typically tasks that they are usually simple that would be probably ninety percent of them that they can be done in hours you're less in terms of our time the larger task typically have some color grant funding attention so in terms of retrieval volume that's not really the issue it's how you get at it if it's something as simple as like email diabetics that's great oh now giving the diabetics that we've seen on this day in this day had this procedure did this we're excluded by that that's where it tends to take up more time you're my sources the throughput is not the issue I pull gigabytes of Dennis and data provided the system self-check is it is the native proprietary that health care system or do other schools across campus have access to that day in order to do said they do essentially we have babies arima the tournament at UNC where if your skills researchers Public Health anywhere you have them valid IRB with the university or higher the elsewhere that's our universities blessed you submit a project for us and provide the data near costing there is a cost when it exceeds approximately at eight hours of time we do try to recover costs for even the ones in the lower but that's kind of it's a touchy subject we try to accommodate everyone try to definitely compliment students because it all from above it when people down locust I'm assuming there must be some type of very strict guidelines I imagined in their childish touring that are using it because of the possibility of a laptop being stolen they cited an agreement where the data is to be contained it is specified in network storage location okay we're transitioning to a convention of three silos IPS research computing as a secure research work space where's the sandbox you can't take anything out of the softbox put something in to help your analysis of the things that I didn't do and setting us all up was to say that we're going to have a break after everyone speaks and then we're going to do some brainstorming about ways that we can work together

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *