DeltaHealth

• 

  Analyzer page. This is where users input their symptoms and obtain advice.

• 

  Emergency Contacts page.

• 

  Disease Catalogue Page. Showcasing outlines of treatments, outlooks, durations, and so much more for a variety of illnesses.

Inspiration

In response to the growing challenge of long wait times in Canada's healthcare system, our team developed a user-friendly app designed to assist individuals in managing their health while they await professional medical care. DeltaHealth empowers users by providing preliminary insights into their symptoms and offering actionable advice to ease their concerns.

We believe that by bridging the gap between symptom onset and doctor consultation, we can enhance patient experiences and alleviate the stress associated with healthcare delays. Join us as we aim to make healthcare more accessible, informative, and supportive for everyone! 🚀

What it does

DeltaHealth is designed to listen to your symptoms and utilize supervised machine learning to classify illnesses based on the symptoms you provide. It then assesses the severity of your condition and offers guidance on next steps.

This solution is aimed at the vast majority of individuals who may not have immediate access to a doctor, providing instant insights to determine whether a seemingly obscure set of symptoms could indicate a serious or life-threatening condition. 🚑

How we built it

DeltaHealth was developed with a modern tech stack for both the frontend and backend. The frontend was built using Next.js, styled with Tailwind CSS, and leveraged components from shadcn/ui for a polished user interface. On the backend, we integrated a tuned Cohere's Command-R to process and condition user input to be fed into our model. 🤖

For the supervised machine learning component, we implemented three algorithms using the sklearn library: the Random Forest Classifier, the Naïve Bayes Classifier, and the Support Vector Classifier. These algorithms were used to accurately classify diseases. Additionally, the Cohere LLM was employed to generate concise and relevant reports, presenting critical insights to the users. 🤓

Challenges we ran into

The main challenge we ran into was developing a model that could accurately predict the correct illness based on the symptoms. Often we would overfit the data or it would not learn at all due to the nature of the training data we found online. This required hours of searching and training, but we eventually found a set that worked! 🦾

Accomplishments that we're proud of

We are mostly proud of the models we were able to train to detect the illnesses, especially after all those long hours of testing and training.

What we learned

Learning how to make supervised classification models taught us a lot more about NLPs and we can't wait to keep on learning more about how these models are applied to the real world!

What's next for DeltaHealth

DeltaHealth envisions a future where it becomes a support for individuals with chronic illnesses—those who often face limited resources and a lack of community support. Our goal is to transform DeltaHealth into a trusted platform and safe haven where people with chronic conditions can find not only the help they need, but also a sense of belonging. By connecting individuals with similar circumstances, we aim to foster a supportive community that empowers them to navigate their journeys with confidence and hope. DeltaHealth aims to create a world where no one has to face their challenges alone. 💙

Built With

• cohere
• next.js
• pandas
• shadcn
• sklearn
• tailwind