Team 9 (Emir, Shadman)
Brief description:
Our model uses a fine-tuned ResNet18 to predict Alzheimer stages from MRI images with up to %90 percent accuracy.
Current dataset that we are using https://www.kaggle.com/datasets/sachinkumar413/alzheimer-mri-dataset
To use our model: Step 1: Download everything in our repository + dataset from the link. The file structure should be that every category folder of MRI pictures, should be under the main folder "Dataset"
example file path for category subfolders: MAIS202\Brain_Disease_Detection\Dataset\Mild_Demented
Step 2: Using alzheimer_resnet18.ipnyb, train the model on your machine. It will take some time, but it will save the weights for the parameters of the model to "model.pth" so you won't have to train it again.
Step 3: Use run_model.py to upload your saved weights and use the model