Overview ======== Let’s face it, when dealing with segmentation of microscopy data we often do not have time to check that **everything is correct**, because it is a **tedious** and **very time consuming process**. Cell-ACDC comes to the rescue! We combined the currently **best available neural network models** (such as `Segment Anything Model (SAM) `__, `YeaZ `__, `cellpose `__, `StarDist `__, `YeastMate `__, `omnipose `__, `delta `__, `DeepSea `__, etc.) and we complemented them with a **fast and intuitive GUI**. We developed and implemented several smart functionalities such as **real-time continuous tracking**, **automatic propagation** of error correction, and several tools to facilitate manual correction, from simple yet useful **brush** and **eraser** to more complex flood fill (magic wand) and Random Walker segmentation routines. See `the table below <#comparison_table>`_ **how it compares** to other popular tools available (*Table 1 of our* \ `publication `__). .. raw:: html

Feature	Cell-ACDC	YeaZ	Cellpose	YeastMate	DeepCell	PhyloCell	CellProfiler	ImageJ/Fiji	YeastSpotter	YeastNet	MorphoLibJ
Deep-learning segmentation	✅	✅	✅	✅	✅	❌	✅	✅	✅	✅	❌
Traditional segmentation	✅	❌	❌	❌	❌	✅	✅	✅	❌	❌	✅
Tracking	✅	✅	❌	❌	✅	✅	✅	✅	❌	❌	❌
Manual corrections	✅	✅	✅	✅	✅	✅	✅	✅	❌	❌	✅
Automatic real-time tracking	✅	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
Automatic propagation of corrections	✅	❌	❌	❌	❌	✅	❌	❌	❌	❌	❌
Automatic mother-bud pairing	✅	❌	❌	✅	❌	✅	❌	❌	❌	❌	❌
Pedigree annotations	✅	❌	❌	✅	✅	✅	✅	✅	❌	❌	❌
Cell division annotations	✅	❌	❌	❌	❌	✅	✅	✅	❌	❌	❌
Downstream analysis	✅	❌	❌	❌	✅	✅	✅	✅	❌	❌	❌
3D z-stacks	✅	❌	✅	❌	✅	❌	✅	✅	❌	❌	✅
Multiple model organisms	✅	❌	✅	❌	✅	❌	✅	✅	❌	❌	✅
Bio-formats	✅	❌	❌	❌	❌	❌	✅	✅	❌	❌	✅
User manual	✅	✅	✅	✅	✅	❌	✅	✅	✅	✅	✅
Open source	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅	✅
Does not require a licence	✅	✅	✅	✅	✅	❌	✅	✅	✅	✅	✅

| Is it only about segmentation? ------------------------------ Of course not! Cell-ACDC automatically computes **several single-cell numerical features** such as cell area and cell volume, plus the mean, max, median, sum and quantiles of any additional fluorescent channel’s signal. It even performs background correction, to compute the **protein amount and concentration**. You can load and analyse single **2D images**, **3D data** (3D z-stacks or 2D images over time) and even **4D data** (3D z-stacks over time). Finally, we provide Jupyter notebooks to **visualize** and interactively **explore** the data produced. **Do not hesitate to contact me** here on GitHub (by opening an issue) or directly at my email padovaf@tcd.ie for any problem and/or feedback on how to improve the user experience! Bidirectional microscopy shift error correction ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Is every second line in your files from your bidirectional microscopy shifted? Look `here `__ for further information on how to correct your data.