Posted on

Week 7 Group status report

What are the most significant risks that could jeopardize the success of the project?
Since carnival starts next week, our biggest risk is not putting in enough time over the weekend to finish our abnormal heart sound classification for MS/AR.
How are these risks being managed? What contingency plans are ready?
We plan to set up a meeting at least twice over carnival to all work togther on this to ensure we are keeping on track.
Were any changes made to the existing design of the system (requirements, block diagram, system spec, etc)? Why was this change necessary, what costs does the change incur, and how will these costs be mitigated going forward?
No changes made this week.
Provide an updated schedule if changes have occurred.
No changes made this week.
Posted on

Eri: Week 7 Status Report

  • What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week (12+ hours).
    • I found more datasets to train the heart sound algorithm on.
    • I denoised all the data before putting it through the CNN.
    • The accuracy percentage went down when I ran the CNN on the denoised heart, but that is most likely because I ran it on my laptop with a much lower iteration and epoch, since it takes a lot more time to train the CNN on  my laptop.
    • Started to train the abnormal heart sounds for AR/MS.
  • Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
    • On schedule.
  • What deliverables do you hope to complete in the next week?
    • We will be working on our CNN for it to reach an accuracy of 85%.
    • Finish training on abnormal heart sounds for MS/AR.
    • Purchase a ballistic gel to play the heart sound through to ensure our testing environment mimics that of a real heart sound well enough.
Posted on

Ari’s Week 7 Report

  • What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week (12+ hours).
    • This week I worked more on the hardware components and trying to make the signal look like the ones in our testing data. I also worked out a testing plan which involves a double blind between our microphone and our ML’s training data. The logic is that if the algo cannot tell the difference and a person cannot tell the difference (>80%), then the signal is about the same.
    • I also helped out with the integration for our demo, getting all of the separate matlab components to connect with each other and worked on modularizing our code.
    • Additionally, I looked into setting up the matlab code to work standalone with python so that way the stethoscope could work without being restricted to having a computer running matlab on it.
    • I also created a team to-do list with our highest action items based on what we learned during the demo, and from the feedback we were provided.
  • Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
    • Our progress seems to be on track, the next thing we need to do is get the testing working and sort out our process for doing so.
  • What deliverables do you hope to complete in the next week?
    • For next week, I am to have a working testing set up, and want to validate our process. I also would like to connect Eri’s shannon energy denoising algo with Ryan’s ML and hopefully that will make the accuracy higher.
    • We also want to find approx 1000 more sound files so that we can train our ML more.