ClickCell 🦠 📸🧬 - Simple, intuitive and productive

Welcome to ClickCell ! This application is designed to provide a low-entry, simple, and fast solution for classifying single-cell images. While intended for haematological single-cell images, this app can easily adapt to various types of images, making it versatile for different research needs.

🚀 Purpose 🚀 ClickCell aims to assist research groups in the initial prototyping phase by offering an intuitive interface for image classification. With this app, you can quickly classify unlabelled single-cell images simply by viewing them and clicking a button corresponding to the desired class.

🖥️ Features

• Easy to Use: A straightforward interface that makes navigation and classification simple.
• Quick Classification: Classify images in just a few clicks, making your research process faster.
• Versatile: While focused on blood cell images, the app can easily be adjusted for different types of single-cell images.
• Cross-Platform: Works on both macOS and Linux, so more users can access it.
• Prototyping Tool: Great for research groups looking to test out classification methods and gather initial data.

📦 ⬇ Download

1. Clone git repository

   git clone https://github.com/LauraWenderoth/ClickCell.git
   cd ClickCell

2. create conda environment and install requirements

   conda create --name env_clickcell python=3.10
   conda activate env_clickcell
   pip install -r requirements.txt

3. Building the ClickCell application using PyInstaller

   pyinstaller --onefile --hidden-import='PIL._tkinter_finder' --windowed --icon=icon.ico ClickCell.py

• Download example data from kaggle

  python download_example_data.py

📥 Getting Started

1. Open the App: Double-click the application icon in the ClickCell/dist folder to launch it.

   

   or start the app via command line

   cd dist
   ./ClickCell

2. Choose Your Images: This will open a folder where you can select the location of the images you want to classify.

   

3. Classification Process: After selecting the folder with your images, the app will display one image at a time, along with buttons for each class and a back button (which allows you to go back one image). As you classify the images, the app will copy them from the original folder to a new folder called $YOURNAME_sorted, creating subfolders for each class. The aim is to have the original folder empty by the end of the classification process.

   

4. Exiting the App: To exit, click the red button in the top-left corner. When you reopen the app, it will pick up from where you left off, with progress updated to show the remaining unclassified images.

   

🛠️ Requirements

• macOS (version 10.15 or later recommended) or Linux
• A collection of unclassified single-cell images

Thank you for choosing ClickCell for your research! Together, we can simplify the classification process and enhance our understanding of hematopathology. Happy classifying! 🎉