MedicalVault | Devpost

Inspiration

The inspiration for this project started as soon as I started getting into the field of Artificial intelligence and data science, and in the age of deep learning, data becomes an important resource to build smart systems. In several fields, we already see that the amount of data that is required to build competitive systems is easily available for some fields, for example, state-of-the-art large vocabulary speech recognition systems that are available from major players such as Google or Nuance are trained with up to 1 million hours of speech. With such large amounts of data, we are now able to train speech-to-text systems with accuracies of up to 99.7%. This is close to or even exceeds the human performance, given that the system does not need breaks, sleep, or ever gets tired. Similarly There is no doubt that healthcare data can be life savior too, it can revolutionize the healthcare industry as a whole, positively impacting the medical care and assistance provided to the patients. But as we can also see the data in the field of healthcare is very scarce and the medical data problem is far from being solved.

And hence was born MedicalVault - A completely new, non-intrusive way to collect health data from both patients and health service providers.

What it does

MedicalVault, at its core, is a cloud repository for medical and healthcare data associated with certain patients and address the following problems:

The platform will act as a repository and will serve patients by giving them a single place to safely store and view all of their health data.
The platform will also act as a data acquisition & filtering repository for the big data analytics life cycle and also help researchers for clinical trial analytics (for example we can track all the patients for side effects of certain drug or vaccine (in cases such as COVID-19)).
Also, we can start integrating some state of the art Artificial intelligence models to provide inference for medical records that patients have stored on our platform and these AI models can also be improved with the help of this data . So, actually the plan is not just to build a platform, the idea is to build an organization for data driven clinical trials, analytics and drug discovery

How I built it

The project entirely depends on the wonders of cloud facilities powered by Django Framework (Django is a high-level Python Web framework that encourages rapid development and clean, pragmatic design) and is wired to REST API Framework, the project is ready to get deployed on a development server.

Challenges I ran into

It is very complex to deal with all the REST API framework and File management at once.

Accomplishments that I'm proud of

The biggest accomplishment will come when we will be an organization serving the humanity but for now taking the first step towards it feels so good and this weekend is worth more than anything from ideation to prototyping!

What I learned

I learned a lot of Django, web programming and cloud concepts from the different resource available online and also through hit and trial.

What's next for Medical Vault

MedicalVaults's has just been started. After the first beta prototype release, we will be testing it with different audience. From there, our feedback cycle will begin which will surely lead to a better and more finished product! Also, integrating many more AI models to get data inference would be our priority!

Case-study : COVID-19

let us imagine our MedicalVault is up and running and has user count in millions and some how we are able to collect around 2000 Chest X-rays and around 1500 CT Scan with mixed patients that is some has COVID some don't have COVID (However we still now who have or not) then we can using four Transfer Learning algorithms: VGG16, ResNet50, InceptionV3, Xception train a model on these images and prepare a preliminary testing platform for COVID 19 this just one application of MedicalVault! (Actually I am working on this demo also but was unable to complete it as it was too much work in platform!)