ledtrack 0.1.0__tar.gz

ledtrack-0.1.0/PKG-INFO

@@ -0,0 +1,162 @@
+Metadata-Version: 2.4
+Name: ledtrack
+Version: 0.1.0
+Summary: Learning-Estimation-Decision Cell Tracking
+License: MIT
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+Requires-Dist: tqdm==4.67.3
+Requires-Dist: hydra-core==1.3.2
+Requires-Dist: omegaconf==2.3.0
+Requires-Dist: tifffile
+Requires-Dist: scipy==1.15.3
+Requires-Dist: pandas==2.3.3
+Requires-Dist: scikit-image==0.25.2
+Requires-Dist: opencv-python
+Requires-Dist: ortools==9.15.6755
+Requires-Dist: imagecodecs==2025.3.30
+Requires-Dist: ete3==3.1.3
+Requires-Dist: scikit-learn==1.7.2
+Requires-Dist: PyQt5==5.15.11
+Requires-Dist: jupyter
+Requires-Dist: matplotlib
+
+# Learning–Estimation–Decision LED
+*A Self-Supervised Learning–Estimation–Decision Framework for Robust Cell Tracking*
+
+![Python](https://img.shields.io/badge/python-3.11-blue)
+![PyTorch](https://img.shields.io/badge/PyTorch-2.14.0-orange)
+
+---
+
+## 📌 Introduction
+
+a novel **Self-Supervised Learning–Estimation–Decision (LED)** framework for robust cell tracking in time-lapse microscopy sequences.  
+
+<img src="method.png" width="500">
+
+The framework integrates:
+
+- **Learning**: self-supervised representation learning to represent cell movement and division pattens  
+- **Estimation**: posterior linking probability based on Bayesian theorem
+- **Decision**: global optimization to resolve cell associations, divisions, and disappearances  
+
+These designs enable stable tracking for unseen data under vary imaging conditions, dense cell populations, and vary cell types.
+
+<img src="example_lineage_tracks.gif" width="500">
+
+---
+
+## 🧰 Dependencies
+### 1. Conda Environment (Recommended for Windows + NVIDIA GPU)
+Download and install Conda for managing Python environments from [Anaconda](https://www.anaconda.com/products/distribution).
+
+Create your environment:
+```bash
+conda create -n your_env_name python=3.11
+```
+and then activate it:
+```bash
+conda activate your_env_name
+```
+### 2. PyTorch Installation (Deep Learning & Large-Scale Image Processing)
+Ensure you have a compatible NVIDIA driver. 
+```bash
+pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu126
+```
+### 3. Pip Packages
+**Note:** The code was tested on Windows/Linux with NVIDIA GPUs but not Mac.
+#### Clone the Repository:
+```bash
+git clone https://github.com/MingweiMin-Lab/LED.git
+```
+enter the directory
+```bash
+cd LED
+```
+You can install all dependencies using:
+```bash
+pip install -r requirements.txt
+```
+Or, you can install the package directly from PyPI:
+```bash
+pip install ledtrack
+```
+---
+
+## 📁 Data Preparation
+
+Place your data in the following structure:
+
+```text
+data/
+├── img/        # Time-lapse cell images (*.tif)
+│   ├── frame_0001.tif
+│   ├── frame_0002.tif
+│   └── ...
+├── mask/       # Cell segmentation masks (same resolution as images, *.tif)
+│   ├── mask_0001.tif
+│   ├── mask_0002.tif
+│   └── ...
+```
+
+✅ Each image must have a corresponding segmentation mask (with sorted name).
+
+---
+
+## ⚙️ Key Parameters
+
+Important tracking-related parameters are defined in the configuration file (`config/tracker.yaml`).
+
+### Core parameters
+
+| Parameter | Description                                         |
+|----------|-----------------------------------------------------|
+| `max_movemment` | Maximum allowed **pixel** distance for cell linking |
+
+### Optional parameters
+
+| Parameter | Description |
+|----------|-------------|
+| `division_detect` | whether to pre-detect division for subsequent self-supervised training |
+| `run_num` | number of parallel processes |
+| `division` |whether to use division constraints in LP|
+| `jitter_thr` | threshold for jitter correction |
+| `post_pro` | post process including pruning and merging tracks|
+|...|...|
+
+---
+
+## ▶️ Usage Example
+**Note:** verify your data and configurations (`config/tracker.yaml`).
+### run main.py `python ledtrack` or the simple code below 
+```
+from ledtrack.train import train_model as tm
+from ledtrack.predictor import predictor as pr
+from ledtrack.cell_tracking import tracker as ct
+
+tm()
+pr()
+ct()
+```
+### or use jupyter
+Launch jupyter notebook and run demo.ipynb
+```bash
+jupyter notebook
+```
+
+---
+
+## 📊 Output
+
+After execution, results will be saved as:
+
+```text
+results/
+├── track.csv                           # Cell trajectory matrix: frame × cell
+├── CTC format result                   # Cell Tracking Challenge format 
+└── visualization of lineage tree       # Lineage tree
+```
+
+---
+

ledtrack-0.1.0/README.md

@@ -0,0 +1,139 @@
+# Learning–Estimation–Decision LED
+*A Self-Supervised Learning–Estimation–Decision Framework for Robust Cell Tracking*
+
+![Python](https://img.shields.io/badge/python-3.11-blue)
+![PyTorch](https://img.shields.io/badge/PyTorch-2.14.0-orange)
+
+---
+
+## 📌 Introduction
+
+a novel **Self-Supervised Learning–Estimation–Decision (LED)** framework for robust cell tracking in time-lapse microscopy sequences.  
+
+<img src="method.png" width="500">
+
+The framework integrates:
+
+- **Learning**: self-supervised representation learning to represent cell movement and division pattens  
+- **Estimation**: posterior linking probability based on Bayesian theorem
+- **Decision**: global optimization to resolve cell associations, divisions, and disappearances  
+
+These designs enable stable tracking for unseen data under vary imaging conditions, dense cell populations, and vary cell types.
+
+<img src="example_lineage_tracks.gif" width="500">
+
+---
+
+## 🧰 Dependencies
+### 1. Conda Environment (Recommended for Windows + NVIDIA GPU)
+Download and install Conda for managing Python environments from [Anaconda](https://www.anaconda.com/products/distribution).
+
+Create your environment:
+```bash
+conda create -n your_env_name python=3.11
+```
+and then activate it:
+```bash
+conda activate your_env_name
+```
+### 2. PyTorch Installation (Deep Learning & Large-Scale Image Processing)
+Ensure you have a compatible NVIDIA driver. 
+```bash
+pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu126
+```
+### 3. Pip Packages
+**Note:** The code was tested on Windows/Linux with NVIDIA GPUs but not Mac.
+#### Clone the Repository:
+```bash
+git clone https://github.com/MingweiMin-Lab/LED.git
+```
+enter the directory
+```bash
+cd LED
+```
+You can install all dependencies using:
+```bash
+pip install -r requirements.txt
+```
+Or, you can install the package directly from PyPI:
+```bash
+pip install ledtrack
+```
+---
+
+## 📁 Data Preparation
+
+Place your data in the following structure:
+
+```text
+data/
+├── img/        # Time-lapse cell images (*.tif)
+│   ├── frame_0001.tif
+│   ├── frame_0002.tif
+│   └── ...
+├── mask/       # Cell segmentation masks (same resolution as images, *.tif)
+│   ├── mask_0001.tif
+│   ├── mask_0002.tif
+│   └── ...
+```
+
+✅ Each image must have a corresponding segmentation mask (with sorted name).
+
+---
+
+## ⚙️ Key Parameters
+
+Important tracking-related parameters are defined in the configuration file (`config/tracker.yaml`).
+
+### Core parameters
+
+| Parameter | Description                                         |
+|----------|-----------------------------------------------------|
+| `max_movemment` | Maximum allowed **pixel** distance for cell linking |
+
+### Optional parameters
+
+| Parameter | Description |
+|----------|-------------|
+| `division_detect` | whether to pre-detect division for subsequent self-supervised training |
+| `run_num` | number of parallel processes |
+| `division` |whether to use division constraints in LP|
+| `jitter_thr` | threshold for jitter correction |
+| `post_pro` | post process including pruning and merging tracks|
+|...|...|
+
+---
+
+## ▶️ Usage Example
+**Note:** verify your data and configurations (`config/tracker.yaml`).
+### run main.py `python ledtrack` or the simple code below 
+```
+from ledtrack.train import train_model as tm
+from ledtrack.predictor import predictor as pr
+from ledtrack.cell_tracking import tracker as ct
+
+tm()
+pr()
+ct()
+```
+### or use jupyter
+Launch jupyter notebook and run demo.ipynb
+```bash
+jupyter notebook
+```
+
+---
+
+## 📊 Output
+
+After execution, results will be saved as:
+
+```text
+results/
+├── track.csv                           # Cell trajectory matrix: frame × cell
+├── CTC format result                   # Cell Tracking Challenge format 
+└── visualization of lineage tree       # Lineage tree
+```
+
+---
+

ledtrack-0.1.0/pyproject.toml

@@ -0,0 +1,37 @@
+[build-system]
+requires = ["setuptools>=61.0", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "ledtrack"
+version = "0.1.0"
+description = "Learning-Estimation-Decision Cell Tracking"
+readme = "README.md"
+license = {text = "MIT"}
+requires-python = ">=3.11"
+dependencies = [
+    "tqdm==4.67.3",
+    "hydra-core==1.3.2",
+    "omegaconf==2.3.0",
+    "tifffile",
+    "scipy==1.15.3",
+    "pandas==2.3.3",
+    "scikit-image==0.25.2",
+    "opencv-python",
+    "ortools==9.15.6755",
+    "imagecodecs==2025.3.30",
+    "ete3==3.1.3",
+    "scikit-learn==1.7.2",
+    "PyQt5==5.15.11",
+    "jupyter",
+    "matplotlib",
+]
+
+[project.scripts]
+ledtrack = "ledtrack.main:main"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.setuptools.package-data]
+ledtrack = ["config/*", "data/**/*", "*.png", "*.gif"]

ledtrack-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+