ConductVision Organoid Monitor

Follow each organoid through fusion and fragmentation

In a pilot, ConductVision follows each organoid under its own ID and logs each fusion, fragmentation or collapse with a video clip.

Plan a pilotWhat a pilot tracks
Two microscope frames of the same organoid culture at 0 and 92 hours, with each organoid outlined and numbered and the same numbers on the same organoids as they grow.
Matthews et al. 2022 · Fig. 2d, cropped · CC BY 4.0OrganoID, a published tool from another research group, outlined each organoid in this microscope time-lapse and kept its number from 0 to 92 hours.

Measures

What a pilot tracks

Identity

  • An ID for every visible organoid, built from its plate, well and organoid number
  • Position, outline and confidence in every frame

Movement

  • Position, distance and speed
  • Direction, rotation and drift
  • Movement relative to neighbors

Before measuring movement, a pilot aligns the frames so camera and plate motion do not count as organoid movement.

Contact

  • Distance to the nearest neighbor
  • Contact events and how long they last
  • Clustering and aggregates

Fusion

  • When two organoids become one structure

Fragmentation

  • Splitting, shedding and breakup

Structural collapse

  • An abrupt loss of area, roundness or boundary
  • Sudden flattening, texture change or disappearance

Each structural collapse comes with its clip and the measurements from before and after it.

Between time points

What happens between time points

Hof and colleagues saw organoids migrate, rotate and fuse with neighbors over up to 7 days of imaging.1 Images taken days apart miss those events and the order they came in.

Over a four-day microscope time-lapse, OrganoID correctly matched over 89% of the organoids it identified from one time step to the next.2 Your pilot checks ConductVision's own tracking and reports a confidence for every track.

Four fluorescence images, 2 hours apart from 80 to 86 hours, of two adjacent organoids merging into one.
Hof et al. 2021 · Fig. 2, cropped, rearranged · CC BY 4.0Two human bile duct cancer organoids fusing between 80 and 86 hours, imaged by light-sheet fluorescence microscopy. Magenta marks chromatin, green marks actin, and the scale bar is 50 µm. In visible light, a pilot's cameras record a fusion like this as two outlines becoming one.

Lineage

Fusion and fragmentation keep their history

When two organoids fuse, the merged structure gets a new ID that records both parents. When an organoid breaks apart, each fragment records the organoid it came from. The original IDs stay in the data, with their measurements from before the event.

Unusual events

Unusual events, saved with clips

A pilot flags a change that departs from the organoid's own history, its neighbors, matched controls or the rest of the experiment. Each flag links to its video clip. When an event starts, such as an organoid fragmenting, the camera images that well more often until it ends.

Camera specification

The camera tracking needs

Tracking needs every organoid in view at every time point.

Field of view

  • Each whole well, in one image or in tiles joined into one

With the whole well in view, an organoid that moves stays in the record.

Timing

  • A full stack of every well every 5 to 15 minutes
  • An image a minute of a well while an event lasts, without delaying the other wells

OrganoID tracked organoids in images taken every 2 hours, and its tracking accuracy held when frames were dropped to simulate a 12-hour interval.2 So tracking does not set the interval, and a pilot images every 5 to 15 minutes to date events and responses closely.

Deliverables

What the pilot delivers

  • An ID and a track for every visible organoid
  • A record of every fusion and fragmentation, with parent and new IDs
  • A time-stamped event log
  • A video clip for every event

Questions labs ask

Does a structural collapse mean the organoid died?

Not necessarily. The camera sees the change in shape. Whether the organoid died needs your own assay.

References

  1. Hof L, Moreth T, Koch M, et al. (2021). Long-term live imaging and multiscale analysis identify heterogeneity and core principles of epithelial organoid morphogenesis. BMC Biology.
  2. Matthews JM, Schuster B, Kashaf SS, et al. (2022). OrganoID: A versatile deep learning platform for tracking and analysis of single-organoid dynamics. PLOS Computational Biology.

Track organoid events in a pilot

Tell us your organoid type, your plates and the events you want caught.

Plan a pilotResearch and accuracy