CiliaDetect -- Deprecated - look for CiliaDetect2.0 == CiliaGUI ----------

Use https://github.com/pmgarderes/CiliaDetect2.0 instead!

Version 1.0 (legacy) A semi-automated tool for fast detection and quantification of cilia from microscopy image stacks (ND2 format)

Authors:
Pierre-Marie Gardères (pmgarderes at gmail dotcom)

• ChatGPT assistance (OpenAI)

---

🚀 Overview

CiliaDetect is a MATLAB-based tool that helps detect, segment, and quantify cilia across multi-channel microscopy datasets.
It supports ND2 files (and downsampled stacks), semi-automated ROI detection, fluorescence quantification across multiple channels, and result export.

---

📋 Installation

1. Install MATLAB (tested on R2021b and newer).

2. Clone or download this repository:

   git clone https://github.com/yourusername/CiliaDetect.git

3. Add the repository and subfolders to your MATLAB path:

   addpath(genpath('CiliaDetect'))

4. Ensure you have the basic MATLAB toolboxes (Image Processing Toolbox).

---

📦 How it Works

The main workflow is handled through a single wrapper file:
▶️ Wrapper_Cilia_Process.m

open the wrapper file (double click on the script in matlab or windows)
set parameters ( change values of the parameters in the wrapper file evaluate line-by-line chunks of code (with F9 or right-click → Evaluate Section).

---

💪 Usage Procedure

1. Prepare Downsampled Data (optional but recommended if you have large files)

If you have many .nd2 files, you can pre-convert them into lightweight .mat stacks:

foldername = 'D:\Path\To\Your\Data';
DSfactor = 25; % Downsampling factor for Z-averaging (default: 25).
batch_downsample_nd2_folder(foldername, DSfactor, true);

This saves computation time later.

---

2. Run the Main Wrapper

Edit and run Wrapper_Cilia_Process.m, which guides you through:

• Selecting the file to analyze (either .nd2 or .mat reduced stack).
• Adjusting parameters for detection and quantification.
• Launching the GUI to detect cilia:
  • Spacebar to detect.
  • Arrow keys to navigate.
  • u to undo last detection.
• Saving the ROIs automatically.
• Visualizing the detected cilia and background masks.
• Quantifying fluorescence across all channels.
• Saving results into an Excel file (.xlsx) for further statistical analysis.

All steps are performed sequentially by evaluating blocks of code.

---

🖥️ Main Parameters (in Wrapper)

• params.load_reduced: 1 to load a reduced .mat file, 0 to process raw .nd2.
• params.DSfactor: Downsampling factor for Z-averaging (default: 25).
• params.windowSize: Size of the detection ROI window.
• params.minArea / maxArea: Minimum and maximum ROI areas.
• params.minElongation: Minimum elongation (for filtering shapes).
• params.minThinness: Minimum thinness (skeleton-based filter).
• params.adaptiveSensitivity: Adaptive threshold sensitivity (0.3-0.7).
• params.backgroundSpread: Dilation size for background calculation.
• params.fluorescenceMode: 'mean' or 'sum' (choose fluorescence computation mode).

---

🌟 Outputs

Each analyzed file will generate:

• .mat files with saved detected ROIs.
• An .xlsx file summarizing:
  • ROI fluorescence (corrected for background).
  • ROI areas and background areas.
  • Multi-channel quantifications.

Example output table:

Cilia ID	Channel 1 Corrected	Channel 2 Corrected	Area	Background Area	...

---

📜 Example: Quick Start

%% Batch prepare ND2 files
dsfactor = 25;
foldername = 'D:\Path\To\Your\Data';
batch_downsample_nd2_folder(foldername, dsfactor, true);

%% Process a File
filename = 'D:\Path\To\Your\Data\reduced_stack\example_reduced.mat';

% Set parameters
params.load_reduced = 1;
params.DSfactor = 25;
params.reload_previous_Detection = 0;
params.windowSize = 100;
params.minArea = 10;
params.maxArea = 1500;
params.minElongation = 2.0;
params.minThinness = 2.0;
params.adaptiveSensitivity = 0.4;
params.maxroiOverlap = 0.8;
params.backgroundSpread = 10;
params.fluorescenceMode = 'sum';

% Load stack and previous detections (if any)
addpath(genpath('.\'));
if params.load_reduced == 0
    [imgStack, metadata] = load_nd2_image_downsampled(filename, params);
else
    load(filename);
end
uniqueDetections = [];

% Launch GUI
view_nd2_with_cilia_gui(imgStack, params, uniqueDetections);

% Deduplicate and save
uniqueDetections = deduplicate_cilia_detections2(ciliaDetections, params.maxroiOverlap);
save_cilia_detections(filename, ciliaDetections, uniqueDetections);

% Visualize
visualize_cilia_masks(imgStack, uniqueDetections, params);

% Quantify and export
results = quantify_cilia_fluorescence2(imgStack, uniqueDetections, params);
resultsTable = struct2table(results);
outputFilename = fullfile([fileparts(filename), filesep, 'MatlabQuantif', filesep, baseName 'cilia_quantification_results.xlsx']);
writetable(resultsTable, outputFilename);

---

📷 Future Improvements

• Full 3D cilia detection across Z.
• Fully automated batch processing.
• Parallel computing acceleration.

---

🎨 Example GUI Screenshot

(Add screenshot here showing detected cilia overlays in different colors)

---

📝 License

This project is licensed under the [MIT License]. see License.txx

---

📝 Citation

see CITATION FILE

---