|
--- |
|
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](https://huggingface.co/antrikxh/Efficient-AD). You can seamlessly integrate this model into your projects. |
|
|
|
# Environmental Impact |
|
|
|
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). |
|
|
|
- **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. 🌟 |
