Inspiration
I was inspired to create this platform when I recently undergone surgery due to appendicitis. This was due to a glaring stomach pain issue, alongside symptoms of nausea and dizziness I was experiencing. However, I had a previous visit to the hospital where I experienced the same symptoms a month prior, but I went home after my stomach felt better but not realizing at the time the type of risk it would put me under. Due to my lack of care for these symptoms, I would soon end up having to face consequences as a result. Medical non-adherences is a common problem amongst many people in Canada. This is due to a lack of understanding the severity of symptoms or not keeping up with medical protocols whether it be neglect, blind advice, self-diagnosing, etc. This is why I designed MedRad as a solution to tracking your medication needs and/or understanding the severity of your symptoms and ensuring that health is at the forefront of one's wellbeing.
What it does
This webapp grants the user a variety of options: Need to track your medication? MedRad's got you covered. Need to understand the severity of your symptoms and what it could entail? MedRad's also got you covered on that end.
To track medication, I set up multiple input boxes that allows the user to enter their medication, dosage, dropdown box of frequency and dosage time. This gets saved into a database that the user can view to keep track of all medications necessary to stay organized.
To check for symptoms, I leveraged Google Gemini's API (Google Search system) to allow for user inputted-prompts regarding their symptoms and analyzing the symptoms via Google Gemini and coming up with an analysis. The analysis includes the user's medical issue, recommendation of over-the-counter medications to take, and procedural actions to take based on user input. Of course, this result is not meant to dictate a medical diagnosis, it is intended to give the user an idea of what they could potentially be suffering through if they don't have immediate access seeing a medical professional. To take things a step further, if the symptoms are potentially severe and life-threatening, There is also include an 'emergency' section, notifying the user that they may be in serious harm and should consider seeing a medical professional and locate the user to two nearby hospitals, if the user wishes to accept sharing their location to the website.
How I built it
Frontend: HTML5, CSS3, JavaScript Backend: Node.js, Google Gemini API
Challenges I ran into
To be honest, I frankly ran into a LOT of challenges. One of the main challenges was trying to integrate Google Gemini's API system to provide accurate results on a user's diagnosis. For example, when I would type in "stomach bleeding", the analysis result would recommend over-the-counter medication like Ibuprofen and would not declare it as an emergency, which is obviously NOT true. After multiple hours, I got the API to work "reasonably" well.
Accomplishments that I am proud of
I was originally supposed to come into this hackathon with a teammate, but unfortunately he couldn't show up so I ended up going solo for this one. Is it perfect? Absolutely not it's filled with issues and contains unfinished parts of the website. Despite the countless debugging errors, the strenuous front-end design, and a one man job, I was shockingly surprised I was able to construct a functioning full-stack platform. I am honestly proud I was able to create something that functions decently well.
What I learned
That nothing is impossible, even if it seems like it. This is my second hackathon I've attended and last time, my code didn't even function. However, this time I was not only able to get the core functionality of the platform to function, but also have time to design the aesthetics of the website. However, I learnt to manage my time properly and focus on the key fundamentals of my project first like functionality and then focus on the latter parts of the project like design.
KEY NOTE
TO ACCESS MY PROJECT, CLICK ON THE MEDRAD FOLDER
Built With
- css
- google-gemini
- html
- javascript
- json
- machine-learning
- node.js
Log in or sign up for Devpost to join the conversation.