🧠 SicTTA: Single Image Continual Test-Time Adaptation for Medical Image Segmentation

This repository contains the official PyTorch implementation of:

SicTTA: Single Image Continual Test-Time Adaptation for Medical Image Segmentation

📝 Under Review / To appear

SicTTA is a continual test-time adaptation (TTA) framework that adapts segmentation models to distribution shifts using only a single test image at a time, without access to the source data or large target batches. It is specifically designed for medical image segmentation where robust deployment is critical.

We also provide re-implementations of several state-of-the-art (SOTA) TTA methods for fair comparison.

---

📌 Highlights

• 🔁 Single Image Test-Time Adaptation (SicTTA)
• 🌊 Continual adaptation across non-stationary test streams
• ❌ Source-Free: no source data or labels required
• 🧪 Comprehensive benchmark on M&MS dataset
• 🧩 Includes several SOTA methods: TENT, MEANT, SAR, SITTA, etc.

---

📦 Installation

# Create a conda environment
conda create -n sictta python=3.10
conda activate sictta

# Install dependencies
pip install -r requirements.txt

---

📂 Dataset: M&MS

We use the M&MS dataset for cardiac MRI segmentation.

📥 Download the dataset and place it in the expected folder.

---

🚀 Source Model Training

To train a UNet segmentation model on the source domain:

python train_source.py --cfg cfgs/mms/source.yaml

• Checkpoints are saved to: save_model/

---

🧪 Test-Time Adaptation

We provide evaluation scripts for SicTTA and other methods. All experiments are configured via YAML files.

# Baseline: No adaptation
python test_time_adaptation.py --cfg cfgs/mms/norm.yaml

# TENT
python test_time_adaptation.py --cfg cfgs/mms/tent.yaml

# MEANT (Mean Teacher)
python test_time_adaptation.py --cfg cfgs/mms/meant.yaml

# SAR
python test_time_adaptation.py --cfg cfgs/mms/sar.yaml

# SITTA
python test_time_adaptation.py --cfg cfgs/mms/sitta.yaml

# SicTTA (our method)
python test_time_adaptation.py --cfg cfgs/mms/sictta.yaml

---

🙋 Acknowledgements

This repo builds upon ideas from:

• TENT
• and others, re-implemented for medical segmentation tasks.

---

📮 Contact

If you have questions, feel free to open an issue or reach out.

Happy Adapting! 🎯