We submitted our Preliminary Report on Friday of last week. On Tuesday, we shadowed Dr. Cohlen in the NICU since Dr. Vesoulis was unavailable. Although this did not directly relate to our project, it gave us further and better insight towards how the NICU operates and how devices in it need to coexist. This Wednesday, Tahj delivered the presentation. We got useful insight from some of the questions asked on how to proceed. In the next few weeks, we will look closer at possible solutions and develop rationale for each of them.
0 Comments
This past week, we shadowed our client Dr. Vesoulis during his round at the NICU. He was able to show us a lot more than the last time we went and we saw some more of the current technologies in seizure detection (more detailed notes from this meeting are in the notebook). We also received feedback on our notebook and realized we had to do better at keeping it up to date to show Prof. Klaesner our progress. Lastly, we met twice to finish our Preliminary Report and will submit it by noon today. The report includes background, need statement, project scope, design specifications, existing solutions, Gantt chart, and team responsibilities. Our meetings consisted of us confirming with each other our understanding of the topic and sharing information on the articles we had found independently. We also discussed our organization to make sure we were on the same page and how we could most effectively complete assignments and share our progress with Prof. Klaesner.
We resubmitted our project scope on Monday and used the feedback delivered Friday to further develop our design specifications. We addressed concerns for units of "missed seizures" and added more specifications to include in our preliminary report. We also developed an outline for our preliminary report and found the past preliminary report of Group 17, 2017 as they had a similar project. After a meeting with Prof. Klaesner, we discovered we shouldn't actually have access to their work, but we will use it just as a reference for report structure, not content. We also did patent and literature research to determine the best and most novel device innovations to include in the paper. Following this, we split up roles for which pieces of the paper we would individually start on before coming together after the weekend. And we hope to meet with Prof. Klaesner with a mostly developed report before Thursday to go over questions regarding content and format.
This week we visited the NICU at the St. Louis Children’s Hospital with Dr. Amit Mathur. While there, we were able to get a first-hand view of the complex, and somewhat hectic, environment. We now have a better idea of the “alarm fatigue” that Dr. Vesoulis was discussing with us earlier this month caused by the sheer amount of alarms and different sensors that are almost constantly blaring. This system of alarms is something that we must take into careful consideration during the development of our product. Although we did not see any epileptic prone neonates on our visit, we plan to shadow the NICU a couple more times in order to familiarize ourselves with the equipment that is currently used to monitor seizures in babies.
Although we are still unsure if we will be able to work on the hardware component of a seizure prediction system, we are continuing to do extensive research into some of the current algorithms and systems being used. We have revised our project scope and plan to meet with Dr. Vesoulis in the next couple of weeks to 1) get a better idea of our design specifications, 2) get an update on the status of the hardware component and 3) discuss a possible timeline for the next couple of months. As this project will have a large software component, we would also like to become more familiar with data analysis techniques before attempting to examine our data in earnest. We have begun working on our website for our project. We will be using a weebly design site for our webpage. We plan to get the website up and running within the next few weeks. We also have finished the first draft of our project scope and will revise and send it to Dr. Vesoulis before turning in the assignment.
After finishing our scope, we plan to visit the NICU while shadowing either Dr. Vesoulis or Dr. Mathur, so that we can get a better idea of the clinical environment. Following our clinical visit, we would like to meet with Dr. Vesoulis once more and begin to delve to the data we have access to for this project. We have discussed various possible machine learning algorithms that we can use to analyze the data such as Random Forest, Support Vector Machines (SVMs) and Nearest Neighbors, but these are all tentative. Currently, we are using papers to try to evaluate which analysis technique might perform the best and whether we have enough time to attempt multiple techniques. Today, we met with Dr. Vesoulis and discussed the project that he had in mind for us for the year.
|
AuthorWrite something about yourself. No need to be fancy, just an overview. Archives
October 2018
Categories |