MOSAIC – One Microscope to Rule Them All?
MOSAIC is a modular microscope integrating twelve imaging modes with adaptive optics in a single platform. Led by the Wesley R. Legant and Srigokul Upadhyayula labs, with key contributions from Eric Betzig (Betzig lab), the system rethinks multimodal imaging from the ground up.
This post highlights its architecture, application range, and strategic relevance for high-end imaging environments.
1. Technical Architecture: What MOSAIC Actually Combines
MOSAIC combines the following imaging modalities via a reconfigurable optical path:
Lattice light-sheet microscopy (LLSM)
Structured illumination microscopy (SIM & 3D-SIM)
Two-photon excitation (point scanning & Bessel light sheet)
Single particle tracking (SPT)
Widefield & oblique illumination (label-free OI)
Photostimulation & optogenetics
Each modality can be paired with adaptive optics (AO) on both excitation and detection paths. Switching between modes takes less than five seconds, enabled by optical toggles and pre-aligned modules.
Key features:
Compact 1 m³ footprint (including lasers)
Multi-color SLM-based LLS illumination
Real-time AO correction via wavefront sensing
Large FOVs (e.g. 1000 × 750 × 10 µm³) with sub-300 nm resolution
Data output: up to 4 TB/hour, processed with PetaKit5D
2. Strategic Relevance: Imaging Core or Overkill?
MOSAIC addresses a familiar challenge: fragmented imaging infrastructures. Most labs rely on multiple systems to balance resolution, speed, phototoxicity, and depth. MOSAIC aims to consolidatethese needs within a single, modular system.
Strategic implications:
✅ Efficiency – no sample transfer across systems
✅ Correlative imaging – same ROI, multiple modalities
✅ Hardware reuse – lower long-term capital investment
Operational challenges:
Complex setup: 1000+ pages of documentation
Steep learning curve: significant training required
Data intensity: demands high-performance storage & processing
Verdict: A powerful asset for central imaging cores, multi-group research hubs, and AI-driven labs – but likely too complex for smaller teams without dedicated infrastructure.
3. Biological Use Cases: Proof of Platform
MOSAIC has already been applied to a wide range of experimental systems:
Long-term live-cell imaging: 24h LLSM timelapse at 260 nm resolution
4D super-resolution: ER & Golgi dynamics in dividing cells using LLS-SIM
Optogenetics: Real-time activation & monitoring of Rac1-induced ruffling
Tissue imaging: ExLLSM of Alzheimer brain samples (~37 TB dataset)
In vivo: Zebrafish regeneration, Drosophila ORN wiring, mouse cortex dynamics at single-spine resolution
→ Correlative switching between modes is possible mid-experiment – without remounting.
4. Forward View: A Platform for Foundation Imaging?
The developers present MOSAIC as more than a microscope – it's a data engine for multimodal, 4D biological models.
Not just an instrument,
but a platform for algorithmic biology.
"The scope of imaging now exceeds what the human eye can interpret.
Training AI on multi-scale image data becomes the next logical step."
— Stefan Prechtl, CELLIMA Scientific Services
5. CELLIMA Commentary
As a consulting partner for high-content imaging and lab strategy, three takeaways stand out:
Modular AO-integration is no longer optional – it's becoming baseline.
In-system modality switching may redefine how imaging protocols are designed.
MOSAIC is not for every lab – but its architecture reflects where advanced imaging is heading.
Let’s Discuss
Does your lab need a MOSAIC – or a more tailored strategy to get closer?
Further reading:
Fu et al. (2025). A Multimodal Adaptive Optical Microscope For In Vivo Imaging from Molecules to Organisms.
bioRxiv: https://doi.org/10.1101/2025.06.02.657494
CELLIMA Newsletter
Strategic insights on high-content imaging, data-driven workflows, and lab infrastructure.
Next issue – September 2025
→ How Pfizer’s NK1 antagonist went from failed antidepressant to blockbuster antiemetic.
Mike joins me to explore the science and strategy behind the pivot.
