Real-Time Prediction of Alzheimer's Disease Progression Using Multimodal Clinical Data
NeuroCast is an advanced multimodal deep learning system developed to address the critical challenge of managing Alzheimer's disease. The platform predicts the progression of the disease—specifically forecasting the Delta MMSE (Mini-Mental State Examination) score—to indicate cognitive decline or stability over time.
Unlike standard clinical assessments, NeuroCast fuses diverse clinical and genetic biomarkers through a sophisticated deep learning architecture to provide neurologists with personalized progression trajectories.
- Multimodal Fusion Architecture: Integrates diverse datasets including patient demographics (age, gender, education) and specific gene expressions (AQP7, RPS5, CHD2, SNX5, ASS1).
- Advanced Deep Learning Models: Utilizes a combination of CNN and LSTM networks to process complex longitudinal data, optimized with custom loss functions and rigorous dataset merging using Pandas.
- Clinical Dashboard: A responsive, real-time interface for doctors to input patient visit data, track historical MMSE scores, and generate AI-driven clinical trends.
- Delta Prediction: Focuses on predicting the change in cognitive capability (Delta MMSE), providing a clear clinical trend rather than just a static diagnostic classification.
- Seamless API Integration: Connects the interactive Next.js frontend to a dedicated AI backend for real-time inference and database syncing via Prisma.
Frontend & Dashboard:
- Framework: Next.js (React)
- Language: TypeScript
- Styling: Tailwind CSS
- Database ORM: Prisma
AI & Data Science Backend:
- Architecture: Multimodal Deep Learning (CNN & LSTM)
- Data Processing: Python & Pandas
- API: RESTful architecture for model inference
NeuroCast was developed as a comprehensive university graduation project bridging the gap between Artificial Intelligence, Software Engineering, and healthcare. The project was completed under the academic supervision of Dr. Syed.
This is a Next.js project bootstrapped with create-next-app.
Ensure you have one of the following package managers installed: npm, yarn, pnpm, or bun.
- Clone the repository:
npm install
# or yarn install / pnpm install
npx prisma generate
npx prisma db push