Music and Dementia

October 29th 2022

Project Fuxi is a student-led initiative in collaboration with Apex Harmony Lodge and with the Coding For Good service club at my school UWCSEA East Campus, Singapore. The school has worked with Apex Harmony Lodge, Singapore for many years, and is dedicated to improving the lives of the elderly suffering from dementia through music therapy. As a part of our dedication to service and giving back to the community, students from our school would go to the Lodge every Friday after school as part of their service club and work with the patients. Through music therapy, they have been able to alleviate some of the symptoms of patients suffering from dementia and allow them to regain their cognitive functions. However, this process is still extremely labor intensive, requiring caregivers to research which songs are most likely to have an impact on the patients based on their background information. Our school launched the student-led initiative to facilitate the development of a technology-based solution Project Fuxi, to create a scalable solution that is user-friendly and that could be used in retirement homes around the world and by those caring for people with dementia at home. This will require securing rights and access to music libraries as well as extensive development of the platform itself.

We know that if done well this can make a massive impact throughout Singapore and beyond, and we hope to partner with local and international organizations to make this a reality. Together we can provide significant support and relief to a group that really needs our help!

As a chair of Coding for Good I was excited to collaborate with Apex Harmony Lodge and come up with a solution. It was clear that we could exploit the scalability of technology and create an algorithm to find songs faster and more efficiently. Our idea was to allow patients to use a Spotify-esque app that will fine-tune its recommendations based on their reactions to the music that's being played. The app would also generate an initial database of songs that are most likely to invoke a positive reaction in the patient, which they would have listened to the most in the past based on their age, ethnicity, origin, personality, etc. The caregivers will be able to input whether the patient is enjoying the music that's being played to them or whether to skip to the next song. Gradually, once enough data has been collated, the app will use advanced algorithms to give more precise recommendations based on the user's preferences.

Recently, my teammates and I completed the first version of Project Fuxi. It uses React.js and MySQL at the core for the front end and database respectfully and is compatible as both, a native app for iOS/Android and a website for all screen sizes. I created the underlying algorithm for selecting and recommending songs, as well as setting up the database. This system is highly adaptable and can be changed to support various forms of authentication and licensing for music.

In the future, we hope to further incorporate technology by introducing hardware that is able to gauge the facial and physical responses of patients who are enjoying the music potentially through tracking facial expressions or using headsets to analyze brain waves.