Decision Support System for Alzheimer's Detection
A Decision Support System that predicts Alzheimer's Disease and tracks progression over time using deep learning on MRI scans from the ADNI dataset, with a user-friendly interface for healthcare professionals.
Problem statement
In collaboration with a team and advisor, we designed and implemented a full-stack clinical decision support system using Next.js and FastAPI to preprocess 3D MRI brain scans and classify Alzheimer’s disease stages from 2D axial slices, enabling early intervention and treatment of Alzheimer's disease.
Thought Process
we focused on building an end-to-end pipeline that could handle the full diagnostic workflow—from raw MRI data to actionable clinical insights. The Deep Learning model was trained on ADNI's comprehensive dataset of MRI scans, allowing for accurate severity assessment. On the frontend, we prioritized accessibility over complexity: the ability to upload patient data and instantly get diagnostic reports streamlines what would otherwise be a resource-heavy clinical process. By integrating privacy-first data handling and real-time visualization of disease progression trends, the system keeps the focus where it matters—on faster, better-informed decisions for patient care.
Demo and link to live project
The link to our report demonstrates patient data upload workflows, real-time AD detection predictions, disease progression visualization through graphical timelines, and secure reporting for clinical decision-making.
Open live project