license: mit
datasets:
- Falah/Alzheimer_MRI
language:
- en
metrics:
- accuracy
- precision
- recall
- f1
library_name: keras
pipeline_tag: image-classification
tags:
- medical
Efficient-AD: An EfficientNetB0-based CNN Model for Alzheimer's Disease Detection
Hey there! 👋 I am thrilled to introduce Efficient-AD, a cutting-edge Convolutional Neural Network (CNN) model designed specifically for detecting Alzheimer's disease using brain MRI scans. This innovative model is built upon the EfficientNetB0 architecture, showcasing remarkable accuracy and performance.
What Sets Efficient-AD Apart?
Alzheimer's disease poses a significant challenge, impacting cognitive functions and necessitating early detection for effective intervention. Efficient-AD takes a giant leap in Alzheimer's detection, achieving an outstanding accuracy of 99.06%. This model is the result of meticulous research and optimization, addressing the limitations of existing CNN architectures.
Key Features:
- EfficientNetB0 Backbone: Leveraging the power of EfficientNetB0, known for its exceptional balance of accuracy and efficiency.
- Deep Funnel Architecture: We've fine-tuned the architecture with a deep funnel design, enhancing the model's capacity to understand complex patterns within MRI scans.
- Transfer Learning Magic: Efficient-AD is pre-trained on ImageNet, providing a foundation for learning intricate features and fine-tuned for Alzheimer's disease detection.
- ReLU Activation Function: Harnessing the benefits of Rectified Linear Unit (ReLU), ensuring faster convergence and improved gradient flow during training.
Performance Overview:
Efficient-AD showcases superior performance, outperforming other state-of-the-art models like DenseNet121, NASNetMobile, and VGG16. It excels in accuracy, precision, recall, and F1 score, making it a frontrunner in Alzheimer's disease detection.
How to Use Efficient-AD:
Efficient-AD is now available on Hugging Face. You can seamlessly integrate this model into your projects.
Environmental Impact
Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
- Hardware Type: Tesla P-100
- Hours used: 72 hours
- Cloud Provider: Google Cloud Platform
- Compute Region: europe-north1
- Carbon Emitted: 3.78
Feel free to explore and integrate Efficient-AD into your projects. Together, let's make strides in early Alzheimer's detection and contribute to improved healthcare outcomes. 🌟