Have you ever had a skin condition?
Have you had trouble affording a dermatologist?
Have you wondered if you could just get a dermatologist on your phone?
Well, look no further!
Hey! Im Anish Kataria, and I’m Krish Malik, and we’re both sophomores from Dublin High School. Today, we’re going to show you Dermalogix, our innovative skin disease diagnosis app.
The purpose of Dermalogix is simple - we hope to make a cheap and accesible way to diagnose a skin condition using just your smartphone so that anyone can easily get a diagnosis.
85 million Americans are suspected to have skin conditions, yet most of these people have never gotten a diagnosis in their life! Sure, dermatologists exist, but not everyone has access to them, and many cant pay hundreds of dollars for a diagnosis.
Now let’s quickly glance over the technology used in the app. This app was built entirely in Swift, and we used Apple's advanced coreML framework to build the machine learning implementation of our application. By using public datasets of images of skin diseases, we were able to use createML, Apple's ML training framework to test and train our machine learning model. After gathering our data and letting our training go through, we had an effective machine learning model which could accurately classify images of skin conditions. Thank you, and we’re dermalogix.
Then, we used UIKit and Apple's Camera libraries to build the scanning functionalities of our app in order to process and implement our ML model. In order to display the directory of local dermatologists, we used Apple's MapKit API. MapKit allowed us to search through Apple's directory of Dermatologists and display professionals near the user's location. We also used MapKit to generate a map and location pointers for nearby clinics, helping simplify navigation for users. Finally, our User Interface and Design was built entirely using UIKit, a User Interface framework which allows you to add the core elements; buttons, labels, text to your app with ease. Our navigation, storyboards and scenes were all also entirely built using UIKit.
This is where Dermalogix comes in; think you have a weird rash? No worries, simply Scan it and get an accurate diagnosis!
Think you have a melanoma spot? All good, just ScanIt!
Think you have a weird mole which you’ve never seen before? We’ve all been there, just scan it!
We decided to use what we knew about machine learning and computer vision to make what we call a "Virtual Dermatologist" - You can simply take a picture and our custom machine learning algorithm will be able to find your skin disease disease based on our model, in real time with no delay. Isnt that crazy? After your diagnosis, you can look at our information page to discover more information about your disease, such as the symptoms, causes, treatments and more. Our machine learning model is able to diagnose many common skin diseases, meaning a widespread use for billions of people across the world. However, we also understood that in cases of serious diagnosis, a patient should seek certified medical care. To help administer this, we added a directory of local dermatologists allowing our users to easily seek medical professionals and contact them without hassle. But you want to keep track of your skin condition? Awesome, we made a database for you to record your skin conditions and access them in the future, allowing you to keep track of your disease progression easily. Now, why should you pick our project? Its simple; due to the impact; Research from the Science Daily finds that Skin disease is the 4th largest cause of human illness, and more than 2/3rds of those with skin conditions were unaware of their abnormal skin findings! We are solving a noble problem which affects billions of people in a new way, helping people improve their lives and preventing deaths with a simple tool that is accessible to ANYONE around the world. We are truly changing the lives of many people in a revolutionary way.
In the future, our team plans to help improve the accuracy of our model and include more skin conditions, so that anyone and everyone can use Dermalogix for their skin diagnosis. This would mean larger datasets, conducting surveys and research as well as deploying our model for faster response time.