Arandu - Research plug-in for young scientists

Title
Loading...
Us :)
Our persona
Components
Backend code
Positive mockup
Positive study mockup
Negative study mockup
Thank you!
Concept
Methodologically flawed study at the front page of pubmed

Inspiration

After witnessing the devastating effects that medical research misinterpretation and fake news had in our country, Brazil, during the pandemic, it was made clear to us that something needed to be done about the issue of statistical illiteracy to help everyone, in special, researchers earlier on in their carrers to read better articles, to be more confident in their analysis and make fewer mistakes during clinical practice.

What it does

Arandu is an artificial inteligence plug-in that, through deep learning techniques, analyzes the METHODOLOGY and RESULTS sections of medical studies in order to see if the conclusions metrics chosen by the authors are methodologically consistent.

How we built it

We used Co:Here's API on Natural Language Processing to create an artificial inteligence that was fed study methodologies using Co:Here's Playground. We also developed a statistical framework for study analysis, done through a thorough medical literature review. Then, we developed our 2nd AI through Co:Here's API, to analyze the results and metrics. To finalize, the process was integrated using Python 3.9 and the slides and concept art were sketched on Canva

Challenges we ran into

None of our team members had a programming background or had ever coded before, so everything was really new. In special, dealing with deep learning related coding was daunting at first.

Accomplishments that we're proud of

We are specially proud of Arandu because it symbolizes, for us, overcoming hard obstacles in order to build technology that can help develop the medical field, which we are inserted in.

What we learned

Altough coding should not be feared by non-tech professionals, programming is an art that translates ideas into reality.

What's next for Arandu - Research plug-in for young scientists

We hope to develop Arandu into a fully functional web plug-in that can recognize and correlate multiple different study designs and result metrics, while also developing a sturdy content library for quick statistical reference.