Bhess7/orthodontic-classifier

Orthodontic Condition Classifier

A ResNet18-based image classification model trained to detect orthodontic conditions from dental photos.

Model Details

• Architecture: ResNet18
• Input Size: 512x512 RGB images
• Output: 8 orthodontic condition classes
• Test Accuracy: 72.73%

Classes

1. Crossbite
2. Crowding
3. Deepbite
4. No Treatment Needed
5. Open Bite
6. Overbite
7. Spacing
8. Underbite

Usage

import torch
from torchvision import transforms, models
from PIL import Image

# Load model
model = models.resnet18(weights=None)
model.fc = torch.nn.Linear(model.fc.in_features, 8)
state_dict = torch.load("pytorch_model.pth", map_location="cpu")
model.load_state_dict(state_dict)
model.eval()

# Preprocess image
transform = transforms.Compose([
    transforms.Resize((512, 512)),
    transforms.ToTensor(),
    transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])

image = Image.open("dental_photo.jpg").convert("RGB")
input_tensor = transform(image).unsqueeze(0)

# Predict
with torch.no_grad():
    outputs = model(input_tensor)
    probabilities = torch.nn.functional.softmax(outputs, dim=1)
    predicted_class = torch.argmax(probabilities, dim=1).item()

Training Data

Trained on a custom dataset of dental photographs labeled by orthodontic condition.

Limitations

• This model is for screening purposes only and should not replace professional orthodontic evaluation
• Accuracy may vary based on image quality and lighting conditions
• Best results with clear, well-lit frontal photos of teeth

License

MIT License