top of page
Abstract Background

CELL ANALYSIS AI TOOLS

cTrack 

CELL TRACKING & MOTION ANALYSIS

From microscopy time-lapses to per-cell motion data 

​

​

​

​

FROM STATIC IMAGES TO MOVING CELLS

cTrack at a glance 

cTrack is RAYFOS’s automated pipeline for analysing cells in microscopy time-lapses. Given a sequence of microscope frames, it detects every cell, follows each one from frame to frame, and turns that tracking into movement trajectories and per-cell motion data — so instead of a static image, you get a quantitative picture of how cells move and behave over time. What normally takes hours of manual frame-by-frame measurement, cTrack does automatically. 

HOW IT WORKS

Inside the cTrack pipeline

Imaging. The input is a phase-contrast microscopy sequence captured on a Leica DMI8 Thunder scanning system with a 10× objective and high-speed autofocus. These are wide field-of-view scans — a single image can be very large, on the order of 31,000 × 3,900 pixels.

​

Tiling and preprocessing. An image that size can’t be fed to the model directly, so cTrack first breaks it down. Each wide scan is padded and cut into a grid of smaller tiles — a first pass into 1000 × 1000 tiles, then further into 500 × 500 — giving uniform patches the model can handle. Each tile is processed independently and then stitched back into the full frame, so the whole field of view is analysed without losing detail or the spatial relationships between cells.

​

Label-free cell segmentation. cTrack uses a label-free model — it works directly on the phase-contrast image, with no fluorescent stains or dyes required. For each tile it produces a coarse segmentation mask that separates cells from background: enough to reliably locate and distinguish individual cells for tracking, rather than a fine pixel-level outline. Stitched together, these masks give a whole-frame map of every cell that the tracking stage builds on.

​

From masks to tracking. These per-frame masks are what the rest of the pipeline builds on: cTrack then follows the segmented cells from frame to frame to produce the trajectories and motion metrics shown below.

cTrack_Figure3.png

Every Cell, Tracked

Still from the tracked time-lapse — a sample of 25 cells is labelled so the tracks stay clear, each carrying a coloured trail tracing its path. The analysis below covers all 55 tracked cells.

STUDY CASE

cTrack in action

These results come from a real cell-viability experiment. From a microscopy time-lapse, cTrack automatically detected and followed every cell across the whole sequence — 55 cells in this dataset — and turned their movement into quantitative motion analysis. Cell speeds are reported in micrometres per minute (µm/min) and displacements in micrometres (µm). Because how a cell moves is closely tied to its health and viability, measuring the speed and distance cells travel over time gives a direct, objective read-out of how the population is behaving. All of the plots below are generated automatically from the tracked data, with no manual measurement. 

cTrack_Figure5.png

Cell Trajectories 

Each cell’s path drawn from a common origin, coloured by total displacement (µm).

cTrack_Figure6.png

Average Population Speed 

Population-averaged speed per frame (µm/min) — the mean speed across all cells over the sequence.

GET STARTED 

Interested in cTrack for your time-lapse data? 

Analyse your data with cTrack 

Tell us about your experience and we will show you what cTrack can measure. 

Start with free segmentation

Try our label-free Cell Segmentation Tool online, free of charge.  

bottom of page