About the Project Inspiration
The idea for this project came from the growing impact of neurological disorders such as Alzheimer’s Disease (AD) and Mild Cognitive Impairment (MCI). Early detection is critical, yet traditional diagnosis depends heavily on expert interpretation of MRI scans, which is time-consuming and not always accessible. We wanted to build a system that could assist in early and accessible diagnosis using AI, especially for areas with limited medical resources.
How We Built It
We developed an AI-driven system combining deep learning and machine learning:
A Convolutional Neural Network (CNN) was used to analyze MRI images and classify them into: Cognitively Normal (CN) MCI AD We also built a RapidMiner pipeline for structured data: Data preprocessing (handling missing values, normalization) Feature selection Classification using Random Forest Evaluation using Cross Validation
The prediction task can be expressed as:
f(x)→{CN,MCI,AD} Challenges Faced
We faced several challenges during development:
Handling inconsistent and noisy data Managing different data types (numerical, categorical, date) Limited features in structured data leading to lower accuracy Debugging ML pipelines and ensuring correct data flow
What We Learned
This project helped us understand:
The importance of data preprocessing and feature selection How CNNs extract hierarchical features from images Evaluation metrics such as accuracy, recall, and AUC Building end-to-end machine learning workflows
Conclusion
This project demonstrates how AI can support early detection of neurological disorders by providing a scalable and efficient solution, assisting healthcare professionals in faster and more accurate decision-making.
Log in or sign up for Devpost to join the conversation.