Really a New Experience

Content:

My First days at the grad school….quite exhilarating to know about the numerous frontiers in the field of Biomedical Informatics.


The lectures in BMI 502 provide us with a gamut of applications which could drive us ahead in this novel field. Introduction lecture by Dr.Kahol gave me an elaborate insight into the computational applications in biomedical sciences. The plan of this course requirement with project, blog participation and creating a website of our own would leave us with an entirely unique experience.

Lecture by Dr.Dinu on EAV’s slightly intimidated me as I have a Non-Computer science background. But various posts on databases by our friends are quite informative and helped me assimilate the concept of databases.

Dr. Robert Greenes’ lecture was an appropriate pill at the right time to me. His emphasis on choosing research interests and his perception on research gave me a better understanding of the approach towards research in this field.

Looking forward to gain a broader exposure in this field….

Posted by
Harsha

Research Methodology Lecture 8/31 - Dr. Greenes

Dr. Greenes lecture today covered a high level discussion on research methodology and was meant to be thought-provoking and forward thinking for the research projects that lie ahead. I believe Dr. Greenes works in the area of where I’m planning on focusing so I could relate to the areas specific to hospital operations and workflows. My research will be based upon the utilization of the EMR and the timeliness of "real time charting" for improving earlier clinical decisions. I have some ideas developing but after listening to the lecture, I need to do more research to see if my idea has been researched extensively already or not. Finding this research will be a challenge. My gut is it hasn’t otherwise we wouldn’t all be struggling with electronic health records but I may be surprised.

Posted by :  Debbie Carter

Showing the 502 students how to post articles

Content:

Test body of text


Posted by Mithra

Lecture 3 Slides Posted

Content:

Today's lecture on Research Methods will be given by Dr Robert Greenes, the Chair of our department. If you would like to know more about Dr Greenes and his research domains here is the link:

http://bmi.asu.edu/directory/page.php?profile=333&cat=professor&dir=Biomedical%20Informatics

The slides for today's lecture have been posted on Blackboard.

Posted by Mithra

Week 1

Content: The professors gave us a broad overview of the field and an interesting prelude to a very challenging field, which is still in its infancy. I definitely see the need for a merger of information technology with medicine in general, but I think that the skills needed for the same have not yet been well defined. Designing a course like the current one is definitely a necessary first step.
I have a background in medicine and epidemiology and had to acquire some technical skills to achieve the purpose of analysis required for making sense of my data. This was a daunting but rewarding task and process led to understand, store, interpret and manipulate my data efficiently.
I think that the field of biomedical informatics needs one to have sufficient working knowledge in the fields of medicine, statistics and technology with a mastery in atleast one.
Dr Dinu's lecture on EAV system was preemptive for me as I did not have a basic knowledge of database management system. The links posted by Dr Kahol on "Introduction to databases" was VERY HELPFUL. The paper by 'Dinu/Nadkarni' is also a very good resourse to understanding the difference between relational database and EAV and the very specific uses of EAV.
Dr Kahol had asked for an example where "computer technology had to become a little less understandable to become more efficient?". I was wondering if Artificial Intelligence can be one such example. Any thoughts??

Posted by Sheetal Shetty

Week One!

Content:

Well here goes, I think we were supposed to have a different entry for each class but to save space I will put both here.

The first class was great in that we were introduced directly into the many different topics that we will be covering.  With general overview, I find it much easier to see what is happening and what is going to happen and therefore no surprises.  Even better was the assignment of our semester long project.  This allows us to have in mind what we will need to work on and it allows for us to think about it all semester long.  This is great compared to other courses when projects are assigned after the fact and methods/details that may have been useful are long forgotten because there was no clear focus.  With an early project topic, it may seem "oh wow" but in a sense there is no surprise later.  Speaking of project topics, I chose topic #4 Social Networks for Clinical Collaboration  (with some urging).  This topic is one of interest since throughout my engineering undergraduate and other volunteer activities I have found that communication is very key to collaboration and getting the desired results.  Social networking seemingly over the past years have had its focused based on networking sites such as Facebook, Twitter, and even this one but there are other professional social networking sites such as LinkIn that can be used to keep in contact with colleagues, other professionals, and even potential employers.  Turning this around to the clinical side, if clinicians were able to socially network and keep in contact with others hundreds of miles away the amount of benefits are many but at the same time there are many obstacles that also exist.  Continuing with networking, personal websites can also be used to establish a network.  I know that the ASU website provided through www.public.asu.edu/~[ASURITE] is a great tool for producing a website, but I just wanted to put it out there that ASU also provides websites through its Google Accounts.  These sites are typically easier to put together and general templates are provided by Google.

ASU Google Sites : http://sites.asu.edu

The second class was present by Dr. Dinu who provided a very informative session on the different types of databases used: conventional relationship vs. Entity-Attribute-Value (EAV).  He explained the many different uses that would require the use of one over the other.  I have had some experience with databases in the past and I found that his lecture helped to consolidate some of the knowledge from the past as well as put a name to the different database types.  His lecture actually made me realize that one of the database organization methods I've attempted to use in the past was actually EAV and it made me realize why it didn't really work work out.

Well on this note, as a former Software Trainer for ASU's Learning Support Services (LSS), I would like to point out that LSS's software training program has a 1 hour workshop on Access that they typically hold during the semester on the Tempe Campus.  Their workshops are designed so that you can be at the workshop with someone walking you through or you can do the workshop individually with the handouts. (I'm in the process of acquiring the handouts from LSS.)

LSS site:  http://www.asu.edu/lss

Posted by Eric

Neuroinformatics

Content:I believe its a novel idea to convey our views and thoughts on through blog.The previous week was intresting.I sincierely feeel that it was a great thought put by Dr.Kanav for assigning projects which has implementations of various methodologies,techniques and knowledge that we gonna learn in coming days.Although my project is based upon designing of EMR which includes genetic information and decision support.,I spend my week learning about Neuroinformatics.I find it very intresting as it not only provides  understanding of various mechanisms and functions of brain using cumputational methods  managing,storing,analysing and sharing complex information through them.There is lot of reaserch going in neuroscience around the world and hence exchange of information becomes imperative and here neuroinformatics plays a major role.
In my undegraduate coure I have studied the physiology and functioning of brain but I still feel there's lot that I can learn.I would be glad if Dr.Kanav will allow  me to change my semester project topic.

Posted by
Ashutosh Singraur

Week 1

Content:

Hello.  The first class of this semester I choose project 3:  Coordinated Large Scale Testing and Evaluation of Medical Devices.  I chose this project because I have a bioengineering background.  I am really excited to learn how to use biomedical informatics methods and apply them to medical device testing.  I am a bit familiar with overall device testing.  Some considerations for the project would be the end users needs and wants.  I am also brainstorming into cost-efficiency and market strategies.  I am sure this type of information processing will not only be helpful for device testing, but also in the development new innovative medical devices.  I am looking forward to working with my peers in this project. 

A presentation on entity attribute value modeling for biomedical databases was given on August 26, by Dr. Dinu.  This was a great introduction presentations and I thought it was very interesting.  I really have no background in databases, except for using Access a couple of times, so this presentation was very informative.

First, Dr. Dinu compared EAV to conventional databases.  He also gave a brief explanation on when it was more efficient to use conventional databases.  Thereafter, the EAV database was explained in more detail, and how it is being used in clinical data.  Some challenges with EAV databases are efficiency, data extraction and vocabulary standards.

I was very interesting in knowing how much more efficient conventional databases are compared to EAV.  I read an article that stated that depending on the query it was range from 3-5 times faster.  A cost effective solution to making EAV queries faster would be purchasing a faster computer. 

Source: Roland, C. S., Nadkarni, P., Marenco, L., Levin, F., Erdos, J., & Miller, P. L. (2000). Exploring Performance Issues for Clinical Databases Organized Using an Entity-Attribute-Value Representation. Journal of the American Medical Informatics Association , 7, 475-487.

Posted by Pierina Ortiz

Application of Computer in Medicine&EAV data

Content:

In the first week of class of Research Methods in BMI, I got the first touch in the Biomedical Informatics field. I learned from the first class that the core idea about the BMI is the application of computer and informatics technology in biology and medical practices, especially the help of computer in clinical application. Dr Kahol gave us the first introduction about the main aspects that computer can be used in medicine, which includes the application of computer in diagnostic support, information management, process manager, education aid, and clinical intervention. Previously, with my background in biomedical engineering, and the short working experience in hospital, I have a general idea about the computer in diagnostic and information management, such as the automatic interpreting of the ECG curve, and the management of the patient billing system and PAC system for medical imaging. But Dr Kahol's introduction extends my understanding of the connect between computer and medical practices, in which there should be more interesting questions and exciting promising future in BMI.

For the EAV database introduced by Dr Dinu, this is a totally new concept for me. I knew the conventional relational database before. But with the comparison of EAV database with the CR database, the EAV database absolutely have some advances in flexibility, and efficiency in dealing with sparse data. However, for query record in EAV database, it seems more complicate than converntional relational database. To use SQL  work on EAV database need writing more than on the CR database. Below is the link for explanation of EAV database system, and one example about how to use SQL to create EAV database system. Hope it can be helpful.

Introduction for EAV database:
http://ycmi.med.yale.edu/nadkarni/eav_CR_contents.htm 

Creation of EAV database:
http://weblogs.sqlteam.com/davidm/articles/12117.aspx
Posted by

Di

Week of 08/24/09 in BMI 502

Content:
Some of the topics covered in the first class were uses of computers in medicine, problems that can occur in research, and how to create a student website on the asu site.  A few of the subjects covered in the computers in medicine subject were data and text mining, social networks, computer-aided medical diagnoses, and managing medical information with computers.  Some of the problems that can occur in research are lacking objectivity in proficency measures and preformance mesures, trouble with adapting to technologies, and having a lack of test subjects.  The student research projects were also covered in the class.

The second class was on databases including Entity-Attibute-Modeling (EAV) and conventional relational databases.  EAV uses only 3 columns.  Conventional relational databases can have a variety of column numbers.  Conventional relational databases can leave a lot of empty data spots.  EAV can leave less empty data spots than conventional relational databases.  How logical and physical structures can be different in databases was taught.  EAV databases can offer more flexability with schema evolutions than conventional relational databases.  The presentation also instructed that a variety of the attributes of each database type should be looked at when someone is choosing a database type to use.

In the first class, I found that a lot of interesting points were covered about computers in medicine.  I liked getting to think about the large number of ways that computers can assist with medical treatments.  I thought about how it is quite a challenge to make computer interfaces intuative to use and powerful in their ability to retrieve relevant information for treatments.  The information in the second class makes me concider that I will try to think about what the data in a database will look like before deciding on a database type to use.  I will also try to remember that while EAVs can avoid empty data spots, sometimes those empty data spots can be important.  The information about databases on the blog can help me to learn more about them and I appreciate that it was posted.

I would recommend MySQL for anyone who wants to learn more about databases.  http://www.mysql.com/ .  MySQL is a program for building and mantaining databases that is able to integrate into a lot of programming languages.  MySQL is open source database software so users can use it for free and donate to it if they want.  Some useful online books on MySQL are availible through our access to SAFARI in our online ASU library cataloges in the research databases section.

Posted by:  Nate

Computer Technology Less Understandable

Content:
I haven't thought about DOS in many years and I dare say, the age of some fellow classmates may have them researching the days of DOS  PCS but I can remember the invention and conversion of DOS to Windows.  Early adopters of computers at home ((for those willing to pay a large sum of money for a home computer)either needed to learn the DOS commands to function or they didn't adopt the technology.  This was probably somewhat of a barrier in the early days of computers but nonetheless, the home computer was nothing short of a miracle.  To not have a use a typewriter any longer was great but DOS based programs were limited and the diversity of what could be done by the average person was minimal.  Then came the first version of Windows.  The user interface was completely different than the black and green DOS screen.  What seemed to be an easier interface, more exciting and definitely more flexible, actually was a much more complex set of programming and less understood.  People had to learn how to support the new application, had to rely on computer repair stores and figure out how to load new programs into the new enviornment.  So, even though the user interface might have been easier, people were struggling with the new concept windows functionality brought and thus it became less understandable but truly much more efficient.  As Windows as evolved over the years, it continues to become less and less understandable, more complex but brings increased efficiency.
Posted by :  Debbie Carter

Research Project # 1

Content:
I chose #1 as my project because I am very involved in the EMR design and development as part of my job as well as creating clinical decision support.  I lack indepth knowledge of genetics and am hopeful to learn more about the HGP and how the future of genetics can impact clnicians decisions and help personalize medicine for patients.  This is a very exciting time to be in the informatics field and a project such as this is something we could realize in the very near future.  Part of the EMR project will need to be the consideration of adoption by physicians.  Will they utilize this knowledge and accept the clinical decision support presented for their patients?  If a genetic rich EMR is to be successful and proven, it would require evidence based research to back up the utilization of genetic information.  On going research and updates would be needed to continue building new decision support which ultimately could cause denial of treatments to some patients versus offering a totally differnt type of therapy to others. Some questions to consider during the project are:  Are physicians ready to break away from a known standard of treatment and try / risk something new?  Are patients willing to take the risk?  Will the patient have the ability to opt out of such treatment or request it?  When do you implement it?  How much genetic information is enough to begin a project?  If you don't have all the genes identified, is there a risk of omission?  Would the computer be blamed for not catching something?  The list can go on but we can't let perfection stand in the way of progress.  Healthcare is a complex industry and this project will explore the decisions being made to move forward with new ideas and technology.

It is very interesting to see the class dynamics and diversity of backgrounds and knowledge.  Developing an EMR that includes genetics seems very doable with the knowledge of fellow classmates and it would be a great opportunity to work on the project as a team versus individual.  Do we have this ability or is this project truly an individual effort?

Posted by :  Debbie Carter

Content:
Week 1:

Lecture 1:

The first lecture was all about the introduction to the course. The class was started giving the course description about what biomedical informatics method is. Following that, objective of the course was mentioned and then the summary of the things that are going to be covered in this course, and finally the term project assignment.

Lecture 2:

Dr. Dinu gave the guest lecture on EAV (entity-attribute-value) modeling from biomedical databases. The lecture covered the basic concept of database, approaches of storing and using data from database. Why we need EAV model was covered next followed by its requirement and uses. as its requirement, and its uses. We also studied the physical and the logical schema, metadata, pivoting EAV data. Examples of conventional clinical data and EAV clinical data helped to understand the difference between the two (former one stores attributes as a separate column and each row thus becomes a set of facts; whereas using EAV helps to store a single fact in a row. (ref: nadkarni)).

By "entity", we mean real world objects, that can be a computer, a patient or a student. Each entity has some
features, or properties, which define them. For example, a computer's features may be its brand, date of purchase, price, warranty period etc. Similarly if a person has features like high temperature, high glucose level, then we assume that the person is a patient. These features, which help to define an entity, are called "attributes".

In slides 16 and 17, definition of primary key is repeated. I believe that should be only in slide 17. A "primary-key" is used to uniquely identify an entity. And it can be mistakenly understood as an "entity". However, a"primary-key" is not an entity, it is also an attribute. For example, a student has an ID. If we know his/her ID, we can obtain his information from the database. We don't really need his name to do that (we used his ID as his name). But, that particular ID defines a particular student. Hence, student's ID is an attribute.

I refer this site (http://www.thocp.net/software/software_reference/databases.htm) to understand why we need databases. Why not conventional file systems (like Excel sheets).

Posted by

Prabal

Research Project Overview

Content: Project 3 Coordinated Large Scale Testing and Evaluation of Medical Devices

Informatics will play a dominant role in the future of all aspects of medical devices--from conception, design, use, to the evaluation of data.  Today, medical devices permeate so many types of facilities and many sectors of the medical field.  Although the use of informatics is fairly limited today, it should play a role in every step of the process: pre-clinical stages, clinical testing and research, and post-marketing surveillance.  This would ease some of the burden of scientists, statisticians and epidemeologists, because certain statistical analysis and methods are still somewhat antiquated in large-scale medical sectors.  This would, of course, save time and money.

My educational background is in neuroscience and biology.  I have worked in the clinical research field and managed data for a medical device/drug company.  I would appreciate any feedback--especially from those in the computer science field who might have ideas on new technologies that would make the process of storing and evaluating data easier.

Thanks!

Posted by Annie

cool week

Content:
This week is really cool for me, beacause I start my classes in US. And it is also a chanllenge for me to study evrything in English.
I think professor Kahol is a humorous and charming young man. But there are little difficulties for me to understand his English. Whatever, I have to get over it and I want to learn more.
However, the good news is that since I have learn something about database in China, I almost understood what professor Dinu said in that class. I think the computer itself is complicated and not so understandable, but it makes our life more efficient, so people love it.
And about the project, I find the project I chose is really a big project, but I like it.

Posted by Xiaoxiao

Content: Thanks to "Finally figured out!!!" Guess I need to learn blogging and databases this weekend. The first class was interesting. Kanav is certainly entertaining. Obviously, many of us have concerns about the project, but I trust promises will be kept and we'll be guided through the process. EAV's aren't logical for me-they seem redundant and tedious, but it's not about me, is it? Thanks for all the posted references.

Posted by: Lee

Database and Relational Database Tutorials

Content:

Here are some links to tutorials on databases and relational databases. Let us know what you think.

Comprehensive introduction to databases:
http://www.databasejournal.com/sqletc/article.php/1428721/Introduction-to-Databases-for-the-Web-Pt-1.htm

Short (fairly easy to read) introduction to databases:
http://www.digital-web.com/articles/introduction_to_databases/

Introduction to Relational Databases (RDBMS):
http://www.utexas.edu/academic/cit/howto/resources/database/dbdesign_mortals.pdf

Posted by Mithra

Profile Setting

Hi Students,

The account provided to you is common to all students. So do not change any settings on the blog.

Thank you.

- Mithra.

Google Code Jam!

May interest some of you (the operative word being 'may').

- Mithra.

~~~

On your mark, get set... CODE!

Are you energized by cracking conundrums? Are you keen to crank out some code? At Google, we know the rush of encountering a challenge and rising to meet it, transforming a problem into a solution and a solution into code. Since 2003, we've been sharing that experience with a global community of computer scientists through our annual programming competition, Google Code Jam.

We're excited to announce Google Code Jam 2009, powered by Google App Engine. Join the fun and compete in several 2½-hour online rounds, attacking three to four difficult algorithmic problems during each round. You may use your favorite programming languages and tools to code up a solution. When ready, run your solution against our fiendish test data. The algorithm needs to be right, and it needs to be efficient: when N=10000, O(N3) won't cut it!

Think you're up for the challenge? Visit the Google Code Jam site to register, read the rules and — most importantly — begin to practice by trying out the problems from last year's contest, so you'll be in shape when the qualification round starts on September 2. You could be one of the top 25 competitors who will be flown to our Mountain View headquarters to match wits for the $5,000 first prize, and the title of Code Jam champion!

Registration closes on September 3, so sign up now!

Lecture 2 posted

The PDF of Dr Dinu's lecture on EAV Modeling has been posted on Blackboard under "Course Documents"

- Mithra.

Information Super-highway in India

check this out

http://www.business-standard.com/india/news/apollo-begins-project-to-connect-health-players/368246/

What is going to be interesting about this article is that as India is building its unique identification number project (like the social security numbers in US), it is implicitly trying to link medical records with it. fascinating concept and one that may be a game changer....

-
Kanav

the philosophy of EAV versus relational database

Mithra will upload some presentations on relational database which you are expected to go through and understand them....

However one of the interesting factor that the class today discussed was that "what is easy to understand for humans may actually not be efficient for humans".... can you all think of some other examples where computer technology had to become a little less understandable to become more efficient?

-Kanav

Practice before you operate

what pre-operative planning can do

http://dailygleaner.canadaeast.com/liveit/article/771872

The science of Motivation

Motivational interfaces. functional fixedness, and are rewards always the best way forward to increase creativity?
form your own opinion... see http://www.ted.com/talks/dan_pink_on_motivation.html
Kanav
Also think of how technology adoption in clinical world can be increased. Do EMR's today make doctors more clericals and if so can we make them better...

for those more interested in technology adoption look at Rogers, E. M. (1995). The Diffusion of Innovations (4th ed.). New York: Free Press.
-Kanav

TA hour

I will be in the student lounge on the second floor between 10:30 am - 11:30 am. This area is adjacent to the HMS lab (Rm 268).

Please sign your name in your posts. Otherwise we have no idea who the author of the entry is.

Thanks!

- Mithra.

Finally figured out!!!

I was trying to figure out how to post an entry and realized that you need to login as Dr Kahol using Sign In link which is on the upper right hand corner of the webpage and then click on New post link. Posting this for people who have never blogged before (like me).

Creating Websites

Here is a link I found that clarifies some questions regarding registering your personal space with ASU.

http://help.asu.edu/Personal_Websites_FAQs

Once you are ready to upload you files to your space, this link would come in handy.

http://help.asu.edu/Uploading_Your_Personal_Webpage

The extra grade is for creating your personal website. If you have any issues with website creation contact me or Dr Kahol and we can help you out.

Here are a few samples of the ones created last year by the 502 class.

http://www.public.asu.edu/~spinne/bmi/
http://www.public.asu.edu/~mvankipu/
http://www.public.asu.edu/~rpsulli/
http://www.public.asu.edu/~adragoto/
http://www.public.asu.edu/~kmukund/
http://www.public.asu.edu/~wewilkin/

- Mithra.

The first class and the next one.

The learning objectives of this particular class were simple... lets immerse ourselves into the research methods of BMI. projects were assigned and brief overview of research projects was presented.
Students want some examples of good reports for their projects. So here they are.
http://www.jamia.org/cgi/content/abstract/4/3/213

http://www.jamia.org/cgi/content/abstract/10/1/1

http://www.jamia.org/cgi/content/abstract/1/3/207

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1029698

http://www.jamia.org/cgi/content/abstract/7/5/431

Now for the fun part of looking at everybody's summaries.

-Kanav

AMIA Podcasts

http://www.amia.org/content/podcasts

This website is a good source of getting into the thick of things with current topics in informatics. check the one on meaningful use and meaningful uselessness out. Comment on it...

Research Projects

The class will have to complete research projects. The idea behind research projects is that the class will learn a variety of techniques in lectures and their project will involve them thinking about how to apply some of these techniques to solve some clinically relevant problems. The problems of interest in this class are as follows

1. Design an EMR that includes genetic information and decision support. As we move towards the brave new world of clinically relevant genetic information (not any of that finding gene for chocolate liking stuff!), it is important to start envisioning EMR of the future that included genetic information. However the perils are many... doctors practicing may not know enough about genetic information, reliability of information is an issue, knowledge discovery is required and privacy is a concern. This project would look at how you would go about building such an EMR and conducting research on its design, development and evaluation

2. Take-home Medical Education. Medical Education is increasingly becoming a major challenge facing the educational community. This project will focus on how biomedical informatics can help. Specifically the project will focus on what does it take to build an information architecture that supports take-home or offline medical education. Today it is virtually impossible to teach medical students and residents everything in class: not the least of which are technical skills. In this project students will enlist strategies and requirements for supporting take home education. Further they will also define methodologies to test this architecture and its effectiveness in offering robust medical skills and knowledge.

3. Coordinated Large Scale Testing and Evaluation of Medical Devices. Medical devices are increasingly permeating our world. Today blood sugar monitors, EKG monitors are commonly available for at home usage. Other type of medical devices such as cardiac stents are critical and important part of healthcare. Each of these devices need to be tested and often in large scale multi-institution studies. How can informatics help? This project will look at design, development and evaluation of informatics architecture to test medical devices.

4. Social Networks for Clinical Collaboration. Well we use these for a lot of purposes including finding the long lost girlfriend, but can we develop social networks for clinical collaborations. What are the challenges and can we use novel technologies such as SecondLife to do this.

5. Neuroinformatics. Computational neuroscience is a burgeoning field and is producing massive quantities of data. The big question that faces this field now is how can the data that is being produced stored, standardized, shared and used for further knowledge discovery. Looks like a potentially excellent area for BMI students to test their research skills!

Exploring the dark side of research in biomedical informatics

This blog is associated with the course titled Research Methods in Biomedical Informatics taught at Arizona State University. The idea behind this blog is to bring together the students, instructors in publishing their experience in the class and what everybody finds interesting, collecting new and unknown facts about research methods and plainly being proactive in seeking knowledge and making this an interactive experience.

For readers, it is guaranteed fun as you will come to know more about BMI and some of the exciting technologies in BMI. For participants in the class this is going to be fun as they can share their experiences and knowledge.

For the course instructor, Kanav Kahol this is going to be scary as the students may annhilate him... but then again that happens anyway.

Kanav