Inspiration

I grew up in unpleasant environments and learned to cope by listening to certain kinds of music. Over time I realized I could self evaluate my mood based on the music I was listening to. I wanted to make that process algorithmic. I wanted to make something I would use long term and something that could help doctors have an easier time with non compliant patients.

What it does

It measure HR data while a patient listens to music. The music and HR data is then paired and then seeing how listening to certain genres or music can affect mood. Then later known calming music is queued to help with a coming panic attack, anxiety and/or help patients with dementia or Alzheimer become calmer. The music is picked based on learned previous data.

How I built it

I wanted to build it to be very portable so I built it using a Raspberry Pi. I used an ANT+ sensor and Garmin HRM chest strap. The data is received by the raspberry pi and fed into the model which then outputs a song. Everything was programmed in python.

Challenges I ran into

I ended up doing all of the work in 8 hours. I had to travel over the weekend and so I had very limited time to put this together. My Sd card corrupted as I was leaving and I didn't have any team members because I was traveling. They did ask politely, but I had to refuse because I knew I had no set schedule of where I would be or if I would even be near wifi.

Accomplishments that I'm proud of

I am proud of the proof of concept I made that works. I believe it can make a real difference, as I had a troubled upbringing, I understand how important music can be to help people cope with certain situations. This project is just the beginning.

What I learned

I learned a lot about what to do and what not to do when it comes to training data and sets. My major mistake was in regards to training my model, but I believe the foundation is there for me to build on later.