Christine Gerardi

First Two Weeks

For the first two weeks of my research experience, the graduate student with whom I'm collaborating was not here yet, so we couldn't start working on the project directly. However, I was given a list of sub-tasks to work on for the time being. I worked on developing a list of discussion questions for the design workshops that we're planning on conducting. I made separate lists of questions for doctors and patients, focusing on encouraging discussion on what type of application would be practical in a clinical setting, and what would convey information the most clearly to the patients.

I used Google scholar to find a few other papers that might support our research. I found substantial research that supported the idea that a computer interface could be used to help facilitate and focus discussions. Although most examples I found pertained to education, I think that the principle could be applied to medical discussions as well. I also found some actual text-based checklists used to facilitate doctor-patient discussions about various diseases. This makes me think that a dynamic tablet-based checklist would be beneficial.

I was asked to find some API's that might make the implementation of our project easier. Because I was unfamiliar with the concept of the API, I spent some time learning about what they are and their applications. I was then able to find some software solutions that might work for our project.

Because it is likely that we will use the Anroid mobile app platform for our project, I spent some time teaching myself Java, which I had no previous experience with. I found that my background in C and C++ made Java very easy to learn.

Week Three

This week Michael, the graduate student, is here so we can start getting into the project. Michael is working to set up a design workshop with local patients and doctors to find out what kind of tool would be the most beneficial to them. In the meantime, we discussed possible interfaces and the pros and cons of each. I had been looking into the possibility of a spacial interface, but we decided that with current available technology, it isn't really plausible. Also, for such an interface to work, rooms would have to be equipped with projectors and motion sensors, which isn't feasible either. We considered the idea of having information on a monitor with motion sensor input, but decided that that would ultimately be too clumsy to use. We decided the best way to go is with a shared interface on touch-screen tablets running the Android platform.

To get ready to develop on the Android platform, I downloaded the Android SDK and the Eclipse IDE and got started learning how app development works. Using a device emulator on my computer, I was able to get a few simple apps up and running.

Wednesday of this week was Fourth of July, which was a lot of fun. The night before I went to see Spider-man with the other undergraduate researchers, and the next night took part in a campus-wide game of capture the flag. I had a great time running around in the dark with my friends, and got to see some fireworks.

Week Four

This week in the lab, we started working on the design of the application. We brainstormed ideas and worked out how we wanted the basic layout of the application to be. We decided that there would be two interfaces, one that the doctor would use to input information, and one that displayed the information in a clear format that the doctor and patient could review together. I made a prototype application in Android of the patient view that we can show to the doctors and patients when we hold our design workshop.

We also discussed other features to incorporate into the system. We came up with the idea of a print feature, so that the patient could have a copy of all the important information to take home with them. The information could alternately be saved as a pdf and emailed directly to the patient.

This week has been especially busy for me because I started a second class in addition to my full schedule of research. I'm learning a lot, and am managing just fine even though I don't have very much free time.

Week Five

This week in the lab we focused on coming up with different options for our application. When we meet with patients and doctors, we want to be able to present them with many different options for how the interface will be, and for various methods of input. I looked at api solutions for dictation, so that the doctor could dictate information into the mobile device. I also found some sample Android code for a simple drawing application. If we wanted to give our application a function for annotating images, we could refactor this code for that purpose.

We are also looking into social signal processing as a potential tool to improve doctor-patient communication. Social signal processing is the use of computers to understand social interactions through analysis of non-verbal behavior. It often involves multi-modal input. I looked on line for open-source solutions that might support SSP, and found an open framework that allows the recording of social signals through various forms of input and then can recognize the recorded signals. I also found many academic papers that might help us.

This week we had two new graduate students join us in the lab. One of them is from Ireland and one is from Iran, so it was interesting to get to know them. We all went out to lunch together with our professor, and then took them on a tour of campus.

Week Six

This week I submitted my abstract for the undergraduate research symposium here on campus next Friday. At the symposium, I'll be presenting a poster describing what we've been working on this summer and what we hope to accomplish this year.

In the lab, we've been having a little trouble with scheduling issues. We are trying to set up a design workshop with doctors and/or patients to get feedback on what type of features in a communication tool would be most beneficial to them. The initial meeting with doctors will hopefully happen within a week or two.

Another thing that my professor has been having me work on is testing an educational robot for the introductory C/C++ class that she's teaching in the fall. The robot is called the Finch. It has built in object, light, and temperature sensors, as well as motors and a built in accelerometer and LED's. It comes with a set of C++ functions that make it easy to use all the features. To test the robot, I made a number of sample programs utilizing the various sensors.

Week Seven

This week we were able to schedule two interviews with oncologists at a local hospital. In the interviews, we presented the doctors with our ideas and asked them what they thought would be most useful in a software tool to improve patient communication. Both interviews were insightful. One doctor suggested that a website containing the patient's personalized could be auto-generated in conjunction with the tablet interface that we have been looking into. The other doctor also liked the idea of an online information repository, but did not think a touch-screen interface within the context of the meeting would be particularly useful. Another idea that came up was the use of speech-recognition software to provide the patient with an annotated transcript of their discussion.

At the end of the week I presented our work during the poster session at the Undergraduate Research Symposium. It was great to have a chance to talk about my work over the summer to students and faculty members, and to see what my peers have been working on.