amal90888/unet-segmentation-model

🩺 UNet Model for COVID-19 CT Scan Segmentation

📌 Model Overview

This UNet-based segmentation model is designed for automated segmentation of COVID-19 infected lung regions in CT scans. It enhances the classic U-Net with attention mechanisms to improve focus on infected regions.

• Architecture: UNet + Attention Gates
• Dataset: COVID-19 CT scans from Coronacases.org, Radiopaedia.org, and Zenodo Repository
• Task: Image Segmentation (Lung Infection)
• Metrics: Dice Coefficient, IoU, Hausdorff Distance, ASSD

---

📊 Training Details

• Dataset Size: 20 CT scans (512 × 512 × 301 slices)
• Preprocessing:
  • Normalization of pixel intensities [0,1]
  • HU Thresholding: [-1000, 1500]
  • Image resizing to 128 × 128 pixels
  • Binarization of masks (0 = background, 1 = infected regions)
• Augmentation:
  • Rotations: ±5 degrees
  • Elastic transformations, Gaussian blur
  • Brightness/contrast variations
  • Final dataset: 2,252 CT slices
• Training:
  • Optimizer: Adam (learning rate = 1e-4)
  • Loss Function: Weighted BCE-Dice Loss + Surface Loss
  • Batch Size: 16
  • Epochs: 25
  • Training Platform: NVIDIA Tesla T4 (Google Colab Pro)

---

🚀 Model Performance

Metric	Non-Augmented Model	Augmented Model
Dice Coefficient	0.8502	0.8658
IoU (Mean)	0.7445	0.8316
ASSD (Symmetric Distance)	0.3907	0.3888
Hausdorff Distance	8.4853	9.8995
ROC AUC Score	0.91	1.00

📌 Key Findings:
✔ Augmentation improved segmentation accuracy significantly
✔ Attention U-Net outperformed other segmentation models

---

📥 How to Use the Model

1️⃣ Load the Model

TensorFlow/Keras

import os
from huggingface_hub import hf_hub_download
from tensorflow.keras.models import load_model
from keras.saving import register_keras_serializable
import tensorflow.keras.backend as K

# ✅ Set Keras backend (optional)
os.environ["KERAS_BACKEND"] = "jax"

# ✅ Register and define missing functions
@register_keras_serializable()
def dice_coef(y_true, y_pred, smooth=1e-6):
    y_true_f = K.flatten(y_true)
    y_pred_f = K.flatten(y_pred)
    intersection = K.sum(y_true_f * y_pred_f)
    return (2. * intersection + smooth) / (K.sum(y_true_f) + K.sum(y_pred_f) + smooth)

@register_keras_serializable()
def gl_sl(*args, **kwargs):
    pass  # Placeholder function (update if needed)

# ✅ Download the model from HF中国镜像站
model_path = hf_hub_download(repo_id="amal90888/unet-segmentation-model", filename="unet_model.keras")

# ✅ Load the model with registered custom objects
unet = load_model(model_path, custom_objects={"dice_coef": dice_coef, "gl_sl": gl_sl}, compile=False)

# ✅ Recompile with fresh optimizer and correct loss function
from tensorflow.keras.optimizers import Adam
unet.compile(optimizer=Adam(learning_rate=1e-4), loss="binary_crossentropy", metrics=["accuracy", dice_coef])

print("✅ Model loaded and recompiled successfully!")