Inspiration
With his grandma, grandpa, and uncle all having major cardiovascular diseases, one of our teammates was worried that he and his family may be susceptible to these diseases as well. When searching online for a way in which he could accurately determine his chance at getting these diseases, he found the process not only to be non-holistic, but of targeting simply one cardiovascular disease, rather than all cardiovascular diseases in general. The suggestions offered by these sites were either insufficient, only focused on a particular aspect of cardiovascular health, or non-existent. The need for an accurate, all-encompassing prediction system for cardiovascular disease vulnerability had presented itself, and upon presenting this idea to his teammates, they determined that change was paramount. Introducing… Healthical!
What it does
Healthical is a simple, easy-to-use website for people to see an accurate evaluation of their risk of contracting a cardiovascular disease and recommendations for them to reduce this risk. Currently, there are platforms where one can input their diet or their blood pressure and get information and feedback off of it. Our website integrates a variety of information so that all of the data is in one place and the user can get one composite understandable score at the end. Instead of the user navigating across multiple websites and applications, Healthical takes a few minutes to input all the necessary information and produces an accurate and realistic chance of the user having a heart attack in the near future.
How I built it
The Healthical has an intuitive, innovative design with simple input graphics which is user-friendly. The model was constructed through the combination of multiple datasets from a variety of past studies including ones conducted by the American College of Cardiology, the National Institute of Health, the Center for Disease Control, Harvard Medical University, and more. The weightage of each of the factors and the result each piece of information has on the result was determined by the data provided by these medical institutions. Once we constructed the model, we then transferred it into a language the computer would understand, Java. After coding the software on Java, we transferred it over to the backend of a website. On the website, the user, doctor’s office, or lab would input their specific information on the website fields provided and that would run our backend Java program, calculating the percentage chance of having a heart attack. Additionally, recommendations would be provided to help reduce this chance by comparing their data with CDC guidelines for the parameters like diet, exercise, etc.
Challenges I ran into
One of the biggest challenges we ran into was implementing the Java backend aspect of the website. It was complicated as the domain.com website would not work. Then we tried to use google sites to embed our java code inside but only applets would work. Overall, we spent over 4 hours coding in html and relearning the language to try and figure this out. In the end we decided to put a download link to our JAR file as it was the easiest and a free method. Later on when we expand on this further and have more time and money to spend, we will integrate the questions and program into our website rather than having a downloadable file.
Accomplishments that I'm proud of
We were proud of the depth and analysis taken to ensure that the individual was given the most realistic and accurate results. We wanted to guarantee the best data for the individual and with this amazing website, we have the resources to do so. We revolutionized medical access to doctors, but virtually with easy navigation and simple tools. With this, millions of people do not need to worry about medical and transportation costs, which definitely makes a difference in their lives. As these costs take a significant portion, the lives of citizens will come to ease, making this product a multi-faceted solution. Our groups’ biggest goal was to create a service/product that will change the world, and from a bigger perspective, we can say that our goal has been accomplished
What I learned
Our team learned about the variety of factors that play into experiencing an heart attack. We understand the many lifestyle choices we have to make in order to stay healthy. From a project design perspective, our team learned how to integrate different types of data into a version in which someone can understand. The integration of all of this information was done by analyzing many datasets and performing many statistical operations on the data. We learned how to accurately predict the probability of an event happening with these math skills. Looking at the problem through code, the challenge was taking the Java code and adapting it to the html website. I had coded a website previously but had forgotten the language so I had to relearn html and figure out how to create a code wrapper to embed in the google sites. We figured it was easier to try and use domain.com but that for some reason wasn’t working for us so I learned a lot of the ins and outs of Google sites and html.
What's next for Healthical
In the future, we plan to use a neural network to accurately determine the weightage values for each principle factor and their relation to heart disease, as well as a supervised, linear regression machine learning algorithm to strengthen our weightage values for the sub-factors. We also plan to use age, weight, race, sex, height, BMI, and exercise details to predict the blood pressure, resting heart rate, etc. the person with that categorization would have if they have heart disease. We would then ask the patient for his/her actual blood pressure, resting heart rate, etc. and compare that to the predicted values. Using the weightage found initially, we would finally determine a confidence value of how close one is to getting heart disease. This is rather than consolidating all the values into one predictive model. One of our main goals with this product that we have yet to implement specificities in exactly what cardiovascular diseases a patient is more susceptible to. A unique feature we haven’t yet seen implemented and plan to achieve is rather than using generalized cardiovascular disease data, we plan to utilize data from each individual disease to give a personalized confidence value, and then average the confidence values to give a final, generalized susceptibility percentage for cardiovascular diseases in general. If one’s value indicates he/she is susceptible to heart disease, the website would offer tips on what he/she can do to avoid it.
Log in or sign up for Devpost to join the conversation.