Raymond Li – Page 2

Progress Report (IPC)

March 14, 2014March 18, 2014Raymond Li

At the Injury Prevention Center, I have been looking at and analyzing data from a database of pediatric trauma patients from 2007 to 2012. Having received this massive dataset, there was a lot of different routes in which we could have approached this. For the beginning of the semester, I have been organizing data into formats that represented the data shown in an earlier pediatric trauma registry paper. For instance, I would group the patients by year, age, ethnicity, etc. Although I am still in the process of formatting some of the data, such as organizing the data based off of time, I have made progress in creating graphs and visualizations that represented the paper. Moreover, I have played around with other visualization to see whether they are effective in displaying that data, such as the motion graph.
Now that most of the data organization has been completed, my next step is to continue making visualizations based off of my data. After Alvin Chang’s presentation, I started to think about how each of my visualization could tell a story, what audience I wanted to show, and how effectively I could portray this data to them. After completing the preliminary visualizations, I will start to delve into the data and look at any trends or anything of significance that I could potentially visualize. I have sketched a few ideas on what I want to do and hopefully within the next few weeks I will bring them to light. Moreover, within the next week or couple of days, I will start compiling a list of questions that I could ask patients, doctors, and families. Similar to what Alvin did, I would like to be able to embed a story behind the data to make it more powerful. By the end, we hope that we will have a collection of visualization to choose from, and tell impactful stories that promote the cause of injury prevention.

Below are some examples of what I have done so far:

In this graph I compared the number of patients by age group. From this graph we can see that males ages 0-9 usually are the ones who get injured the most. However, this changes the older the children get. It would be interesting to find out why this trend exists.

Update by Jack, March 18th: the interactive image below did not consistently appear in your post, most likely due to Google Drive hosting issues. But I was able to paste the underlying URL into another browser to capture this static image, which I have pasted in. Therefore, both the static and interactive versions may appear in some browsers. See my comment below about hosting code on GitHub instead.

The interactive version:

In the next two graphs, I show the percentages of patients by ethnicity. Each graph represents either male or female patients. Hopefully for future graphs I can mix both male and female so that all the data could be contained in just one graph.

For this graph, I made a times series of female patients based off of ethnicity. I will include the males once I am able to. Here we can see the the number of patients throughout each year. Whites mostly make up most of the injuries, but there was a significant decrease in 2010. I could possibly research why there was such a big drop to see if it was just a coincidence or not.

Note from Jack, March 18th: The Google Motion Chart below appears in some browsers, but not others, which also may be a Google Drive hosting issue. Please try hosting with GitHub to see if that resolves the issue.

Below is a Google Fusion Map that I made which compiles the total number of patients in each city/town from 2007-2012. Around the Hartford area is where the most injuries take place. I would however like to use the Leaflet application instead of the Google Fusion.

Update Assignment 6b

March 6, 2014March 10, 2014Raymond Li

From last week, I have compiled and organized my data so that I was able to create visualizations. So far I have managed to incorporate, age and ethnicity. From these graphs we can see that the majority of injuries occur for males up until they are around age 5-9, however, females are more likely to get injured after 10 years old.

*March 10th update from Jack: Ray is having difficulty with web hosting on Google Drive, as I can only view some of his iframes below in Safari. So I have pasted a static image of two of his interactive charts: a bar chart and a time-motion chart. Looks like hosting on GitHub pages will be a better solution.*

In this graph I compiled a list of patients based on their ethnicity. From the data, whites and males make up most of the injury patients.

Here I created a time series graph that shows the number of patients each year by ethnicity.

I will also be making maps. Although this is a incomplete map, it is a sample that of what I will make when I get the data. For each city, I will include the total population as well as the number of patients in that area from 2007-2012.

Assignment 5 (IPC Update)

March 1, 2014March 6, 2014Raymond Li

In this graph I compiled a list of patients based on their ethnicity. From the data, whites and males make up most of the injury patients.

Here I created a time series graph that shows the number of patients each year by ethnicity.

IPC Google Fusion

February 22, 2014February 22, 2014Raymond Li

For IPC, there are many types of charts and maps that are useful in displaying data. So far I have been moving along with organizing my data. In this map, it shows the number of patients in each residential city/town in 2007. As I progress with the data, I hope to add more information, such as frequency of injury in each city.

Plans for Pediatric Trauma Registry

February 13, 2014February 13, 2014Raymond Li

Over the past two weeks, I have been going to the Injury Prevention Center quite frequently and working with Garry on the data that he gave me. He has given me a published paper on how the IPV has categorized and analyzed the data; therefore I have been using this paper as a guide. The data that has been given to me contains 5 years (2007-2012) of all pediatric traumas that have occurred. The information is extensive and contains information such as age, gender, ethnicity, time, place, injury type, and etc. Because this data base is owned by Connecticut Children’s Medical Center, the data itself is private and cannot go out into the public unless it is shown as a visualization.

The first priority is to organize the data as shown in Using Trauma Registry Data to Guide Injury Prevention Program Activities. By grouping information together such as age and injury severity, we can better see the correlation between different age groups and the result of the injury. By the time I am finished I will hope to have the trauma patients’ demographics, the cause of injury by mean age, temporal characteristics, injury severity, injury location, injury intent and mechanism, and injury location.

Next, as we move along in the class, I will then start making charts that will showcase the information that is interactive but also easy for the public to understand. For instance, I would group together injury types and display which types of injuries are more prevalent, and add other information such as age groups most affected. Here is an example of an interactive chart that could be helpful: Health Chart

As well as injury information, creating maps would also be helpful in looking at cities and locations where injuries often occur. The authors of the Trauma paper, used ArcGIS to map cities in Connecticut, but it would also be beneficial for the map to be interactive as well. Here would be an example: Health Map