Coronary heart disease is really tough to predict with a high amount of accuracy, Owing to its sudden nature, associated with the fact that it doesn’t have any noticeable symptoms.
Risk is defined by the lack of adaptive capacity to the disease, most of the time there is very little time to react, and many often would have to go through a cardiac arrest to find out of their condition.
Hence, my project is really to reduce the immediate risk of the disease, and I addressed it in two ways:
-
Through an accessible detection algorithm to every local Singaporean, this is so they can check if they are facing an immediate risk to the disease. Even if its 1 day or two days before an actual cardiac arrest occurs, that one day would increase what an individual can do given the situation. This will be done through predictions using Machine Learning models on Edge Devices or from the cloud.
-
To more effectively find preventions against CHD, we would need to do more research in the field. CHD is primarily caused by lifestyle, however, there is a lack of research of lifestyle and emotional, mental, physical wellbeing on the development of CHD. This second solution would be to create a publicly accessible dataset that people like us can do machine learning on, and the app, is a very good way to collect data. Crucially, I believe that local, relevant, up-to-date data would trump any overseas data that we can find, because cultural influence would also have an influence on mortality. So we can’t just generalise the datasets from the US to people in our society, without acknowledging that cultural differences are lost in the process.
(MY FIRST HACKATHON!!)
I was actually at a loss of what to do at first. Less than 48 hours, a machine learning based solution? Then I started watching videos on what projects are cool to do in 2021. I landed on Parkinsons disease, which I found after a while a little tough to expand upon and come up with something more substantial, other than its early prediction (which many people have done before). So I chose one that was pretty close to home, because I've had a couple family members suffer from Ischaemic Heart Disease. Ischaemic heart disease was trickier than Parkinsons, because of the lac of noticeable symptoms and its sudden nature.
I also wanted a more local focus, and coronary heart disease (CHD) is also to be a leading cause of death in Singapore. I have a slight interest in medicine and mortality from the books that I read from doctors, and yeah so this was my cause for the machine learning application. I toiled on this over the two days, and now that as I write this reflection I have been awake for 26 hours, and irregardless of the outcome, I feel that this hackathon is a pretty interesting way to throw yourself into the deep end and try to impress yourself by exceeding yourself.
See the video for my app here: https://youtu.be/y4zQZkeHc18
Will probably go for more hackathons in the future. Thank you NTU's EEE for the virgin experience!