import timm
import torch

model = timm.create_model('densenet121', pretrained=True)
model.classifier = torch.nn.Linear(model.classifier.in_features, 14) # 14 bệnh phổi
model.load_state_dict(torch.load('models/weights/chexpert_densenet121.pt'))
model.eval()

def generate_heatmap(image, class_idx):
    output = model(image)
    model.zero_grad()
    output[0, class_idx].backward(retain_graph=True)
    gradients = self.get_activation_gradients()
    activations = self.get_activations().detach()
    weights = torch.mean(gradients, dim=[0, 2, 3])
    cam = torch.zeros(activations.shape[1:], dtype=torch.float32)
    for i, w in enumerate(weights):
        cam += w * activations[0, i, :, :]
    cam = np.maximum(cam.cpu(), 0)
    cam = cam / np.max(cam)
    return self.overlay_heatmap(cam, original_image)

reference_ranges = {
  "hemoglobin": {"unit": "g/dL", "low": 12.0, "high": 16.0},
  "wbc": {"unit": "K/µL", "low": 4.5, "high": 11.0},
  # ...
}

def is_abnormal(lab_name, value):
    if lab_name.lower() in reference_ranges:
        range_info = reference_ranges[lab_name.lower()]
        return value < range_info["low"] or value > range_info["high"]
    return False

cd medical-diagnostic-tool
pip install -r requirements.txt
npm install

cd medical_service
uvicorn main:app --reload

cd ..
npm start

async function analyzeMedicalImage(imageFile) {
  const formData = new FormData();
  formData.append('file', imageFile);
  formData.append('file_type', 'image');
  formData.append('analysis_type', 'comprehensive');
  const response = await fetch('https://your-api-endpoint/analyze', {
    method: 'POST',
    body: formData
  });
  return await response.json();
}