README.md

metadata

license: apache-2.0
datasets:
  - SilpaCS/Augmented_alzheimer
language:
  - en
base_model:
  - google/siglip2-base-patch16-224
pipeline_tag: image-classification
library_name: transformers
tags:
  - Alzheimer
  - Stage-Classifier
  - SigLIP2

Alzheimer-Stage-Classifier

Alzheimer-Stage-Classifier is a multi-class image classification model based on google/siglip2-base-patch16-224, designed to identify stages of Alzheimer’s disease from medical imaging data. This tool can assist in clinical decision support, early diagnosis, and disease progression tracking.

Classification Report:
                  precision    recall  f1-score   support

    MildDemented     0.9634    0.9860    0.9746      8960
ModerateDemented     1.0000    1.0000    1.0000      6464
     NonDemented     0.8920    0.8910    0.8915      9600
VeryMildDemented     0.8904    0.8704    0.8803      8960

        accuracy                         0.9314     33984
       macro avg     0.9364    0.9369    0.9366     33984
    weighted avg     0.9309    0.9314    0.9311     33984

---

Label Classes

The model classifies input images into the following stages of Alzheimer’s disease:

0: MildDemented  
1: ModerateDemented  
2: NonDemented  
3: VeryMildDemented

---

Installation

pip install transformers torch pillow gradio

---

Example Inference Code

import gradio as gr
from transformers import AutoImageProcessor, SiglipForImageClassification
from PIL import Image
import torch

# Load model and processor
model_name = "prithivMLmods/Alzheimer-Stage-Classifier"
model = SiglipForImageClassification.from_pretrained(model_name)
processor = AutoImageProcessor.from_pretrained(model_name)

# ID to label mapping
id2label = {
    "0": "MildDemented",
    "1": "ModerateDemented",
    "2": "NonDemented",
    "3": "VeryMildDemented"
}

def classify_alzheimer_stage(image):
    image = Image.fromarray(image).convert("RGB")
    inputs = processor(images=image, return_tensors="pt")

    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs.logits
        probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist()

    prediction = {id2label[str(i)]: round(probs[i], 3) for i in range(len(probs))}
    return prediction

# Gradio Interface
iface = gr.Interface(
    fn=classify_alzheimer_stage,
    inputs=gr.Image(type="numpy"),
    outputs=gr.Label(num_top_classes=4, label="Alzheimer Stage"),
    title="Alzheimer-Stage-Classifier",
    description="Upload a brain scan image to classify the stage of Alzheimer's: NonDemented, VeryMildDemented, MildDemented, or ModerateDemented."
)

if __name__ == "__main__":
    iface.launch()

---

Applications

• Early Alzheimer’s Screening
• Clinical Diagnosis Support
• Longitudinal Study & Disease Monitoring
• Research on Cognitive Decline