#clear lung project

Currently, the Clear Lung Project Design and Hardware sub teams are collaborating to create the device prototype. It was determined that a vest-like material would be utilized. A weighted vest with a front pocket was purchased and stuffed with a cotton-type material during the testing phases in order to insulate and minimize background noise.  The team plans to attach the piezo microphones directly into the vest where recordings of the lung sounds will take place.  Four piezo microphones will be attached to this vest, with one in each quadrant of the back: upper-left, upper-right, lower-back, and lower-right. At this point in the project design, the sub teams will be working to maximize comfort and efficiency in sound recording.

- John Obeid

I initially joined M-HEAL as a member of the Incubator, a space where we worked in small project groups to mimic the motions of a full project team (needs assessment, ideation, prototyping). This experience was valuable as it taught me the necessary skills to understand the importance of M-HEAL and the work that we as an organization do every week. I’m glad I had a stepping stone before joining the Clear Lung Project because I am now able to better understand why certain decisions are made and how to best proceed to a solution. My time on the Clear Lung Project  as a part of the Machine Learning (ML) sub-team has been rewarding and insightful. As a sophomore, I often don’t fully understand all of the tools that we use to design our program, but I am surrounded by other individuals in my team who constantly help me grow and understand the project material. This year, I’ve gotten to see our project go from just a data set to a functional program that can be used as a basis when we obtain lung sound data. Each week we work collaboratively to add more components to our program and fine tune what we have in hopes to find the ML model that best fits with the data.

I am excited to see where the Clear Lung Project takes us in the future. I hope that one day, all of the work that we have put in and challenges we have overcome reflects in a finalized product that makes an impact in the society around us.

We haven’t had many social events this semester, but I was able to make a get together for pizza. It was nice to get to know some of the members more outside of the project-based meetings. Being on this team, I am able to work with people of different race, religion, gender, age, etc. All of us working together towards a common goal to help others. To me, this is what makes M-HEAL and Michigan so great.

Unfortunately I am a senior and will only have this year to contribute to the Clear Lung Project, but I wanted to find a way to contribute even after I have graduated. I applied to go on the SANA trip to which was accepted. We are going to Guatemala, and I am super excited. By participating in this needs assessment trip and creating a new project team from this trip would allow students after me to help which is a way for me to contribute even after I have graduated.

For the last month, the main focus of the Clear Lung Project has been finalizing and collecting our results to present for the Fall Design Review. We have made considerable progress over the last semester and desired feedback on our project’s design and how we plan to implement the device in the future. We have completed a couple major goals for the semester, and are very excited to report that we now have a device that can successfully, and accurately record sound data from lungs, and our software sub-team has extracted sound data and compiled a feature library for future use. 

These two achievements are possibly our greatest of the semester and it shows that we can now implement pre-clinical testing with our device to collect various sound data and attempt to differentiate between the data by using our early prototype. This is very important in our process of completing an IRB form. By doing various pre-clinical trials with our device, it will be much easier to find sponsors and hopefully gain access to CS Mott hospital in order to do future clinical trials and collect sound data from participating patients. In preparation for this hopeful prospect we have begun writing test protocols for both our prototype and code. 

Since winter break our team has been making major progress towards our goals for the semester. We have divided into subgroups that all specialize in different areas of the project. I think that this has made us more effective. Our meetings on Thursdays are now used to catch each other up on what each sub team has been doing instead of trying to accomplish things in such a short period of time. Another big development that has occurred this semester is funding. We were rewarded the BLI grant, which will help us in numerous ways. It will allow us to buy materials in order to construct a prototype and gives us more flexibility in terms of trying different things.  We have also had social events that have helped us get to know each other better. The new funding and faculty advisors we acquired from last semester has everyone excited about what we can accomplish.

 All of the sub teams have different objectives, but one of the major areas of focus recently has been IRB approval. IRB approval would allow us to collect data from the hospital in order to test and refine our algorithm and prototype. This will be a long process but getting started on it earlier will help us in the future. All in all, the rest of the team and I are very happy about our progress this semester and have high hopes for what’s to come.

Team Pneumonia splits into 4 subteams 

Our team is currently in the incubation stage. This past month, we have applied for multiple grants to help start our device creation. We also managed to secure a lab in the LBME lab. Our lab is equipped with hardware to help with the creation of our prototype. Our team split into multiple sub teams. These teams are software, hardware, research, and outreach. My team, the software team, started to look into feature extraction algorithms to process signals. We also started to compile a list of lung sounds to process with Matlab. We hope to have a full prototype by the end of the semester. Our team is really looking forward to this semester.

Earlier this month, I attended the M-HEAL design review. Not only was this experience valuable to see the cool stuff other M-HEAL teams are doing, but at the same time, it also really helped show where our team could be in the future. The first few presentations were all from teams pretty far ahead of where we are at this point; however, observing the steps they have taken to refine their prototypes could help us decide the path we want to take in building our own prototype. For instance, one common theme within each of these established teams was that they all had PR teams, so eventually down the road we may explore implementing a PR team for our group. After splitting up and going into different rooms, the presentations we listened to were from teams that are closer to the current state of our team, and from these teams we learned of direct steps we could take in the near-future. Mainly, these groups displayed steps to take for our early prototyping phase that we are currently in. Using our experience at the design review, our team discussed several goals for the next semester, deciding that our main goals are to actually build the prototype and collect data on pneumonia coughing sounds to help build an algorithm. Hopefully we will have a prototype to share with everybody by the next design review!

Updates from Team Pneumonia

The last month has been a really exciting time for Team Pneumonia! Last semester we did a ton of research, and decided that doing analysis of pulmonary sounds would be the best way to tackle diagnosing pneumonia. This semester’s primary goal was deciding how to create an algorithm to take all of the aspects of the data and turn it into a diagnosis. With the help of our twelve new members, we were able to determine that a machine learning based classification algorithm would be the best way to create a diagnosis. With this plan in mind, we set out to get help from some faculty members, and figure out how to collect enough data to train our classifier. We have been working closely with Dr. Jenna Wiens and Dr. Alberto Figueroa to help steer our project in the right direction and gain new perspective on how to approach such a complex project. Currently, we are trying to tackle our second goal of collecting data. We have been in contact with several universities about sharing data, while also trying to develop our method of data collection. The next several month are definitely going to be a challenge, but I am very optimistic about what we'll be able to accomplish!