NeuroScreen AI

An interactive AI-powered tool to detect early signs of depression and cognitive decline in minutes.

Comment

Inspiration

We were inspired by the urgent need for accessible, early-stage screening tools for depression and cognitive decline conditions that often go undetected until they significantly impact quality of life. Millions suffer silently, and delays in detection can make recovery harder. Our team wanted to build something lightweight, interactive, and useful for self-screening especially for underserved or hesitant populations. The DETECT track of NeuraVia Hacks was a perfect match for this vision.

What it does

NeuroScreen AI is a web-based screener designed to identify early signs of depression and cognitive decline using a subset of validated psychological assessments. It allows users to: Input their age and name Complete a short version of the PHQ-9 (depression screener) Respond to two targeted questions on memory and attention Receive a result summary and recommendation instantly The tool offers users a private, interactive way to reflect on their mental health and serves as a starting point for deeper assessment.

We developed the project in under 2 hours using a lightweight but powerful tech stack:

Streamlit for the frontend interface Python (Google Colab) for backend and logic Ngrok to generate a public URL for easy sharing Custom logic based on PHQ-9 scoring Two select-box questions to simulate early cognitive screening We used pyngrok and os.system() calls in Colab to host the app live without requiring complex setup or external servers.

Challenges we ran into:

🔒 Configuring and authenticating ngrok for live URL generation inside Google Colab 🧪 Debugging minor syntax issues in Streamlit with multi-line code cells ⏱ Time constraints in creating a usable, user-friendly prototype with complete UI/UX in under 2 hours 🌐 Ensuring real-time interactivity while deploying from a cloud notebook

Accomplishments that we're proud of

🚀 Successfully launched a fully interactive screening app ⏱ Completed ideation-to-deployment within the hackathon time limit 🧠 Combined mental health and cognitive health into one streamlined tool 🔗 Generated a working public link from Colab with no external hosting needed 💬 Wrote and recorded a short walkthrough video to demonstrate use

What we learned

How to use Streamlit + Ngrok for rapid prototyping and sharing How to simplify mental health tools for first-time users How to convert theoretical screening frameworks into real-time digital tools How to work fast and collaboratively under pressure

What's next for NeuroScreen AI

🔍 Expand cognitive screening section with more validated questions 📈 Add a monitoring dashboard for repeated use and symptom tracking 🤖 Integrate AI to adaptively recommend next steps or resources 🌍 Translate interface into multiple languages for accessibility 🧩 Build additional modules under CONNECT and PERSONALIZE tracks for long-term use 🏥 Collaborate with mental health clinics to test feasibility in real settings

Built With

Share this project:

Updates