State-of-the-Art AI for Microscopy

Microscopy Data
Analysis at Scale

Browser-based platform for microscopy data visualization and AI-powered analysis. Access state-of-the-art deep learning models for segmentation, tracking, and quantification of biological data. No coding. No expensive licenses. No fragmented workflows.

5D Data Visualization

Explore multi-dimensional microscopy data with Z-stacks, time-lapses, multi-channel overlays, and tiled imaging.

Visual Workflow Builder

Create complex analysis pipelines with drag-and-drop nodes. No coding required—just connect, configure, and run.

State-of-the-Art Models

Access Cellpose, StarDist, SAM and more. Fine-tune on your data with one click—no coding, no fragmented workflows.

3D Volume View X Y Z 100 µm Time 127 / 360 Channels DAPI Nuclei 70% GFP Membrane 90% mCherry Markers 50% Brightfield Trans Off Dataset Info 1.4 TB Total Size 2048² XY Resolution 120 Z 360 T Depth × Time 5D Dimensions X, Y Spatial (2048 × 2048 px) Z Depth (120 slices, 0.5µm spacing) T Time (360 frames, 5 min intervals) C Channels (4 fluorophores)
5D Data Visualization
Workflow Canvas Run Workflow Load Dataset organoids_4D.zarr 2048×2048×120×360 Denoise N2V • Self-supervised GPU accelerated Segment Cellpose 3.0 cyto3 model Track Ultrack Lineage reconstruction Analyze Morphology + Motility Export to CSV Node Library Load Dataset Denoise Segment Track Analyze Execution Status Load Dataset Complete Denoise Running... 66% • ETA: 2m 34s Segment Queued Track Queued Analyze Queued
Visual Workflow Editor
State-of-the-Art AI Models One platform. All the best models. No coding required. Cellpose Cell & nucleus segmentation v3.0 • 12M+ parameters 5,200+ citations Select StarDist Star-convex object detection v0.8 • Convex shapes 2,800+ citations Select SAM 3 Segment Anything Model Meta AI • Foundation model 10,000+ citations Select Ultrack Multi-hypothesis tracking Latest • Cell lineages NEW Select Fine-tune Cellpose on Your Data No Code Required TRAINING DATA Drop images here 12 images uploaded ✓ ANNOTATIONS 247 labeled cells EPOCHS 50 ~15 min on GPU Training Progress 60% Start Fine-tuning Estimated: 15 minutes Why This Matters No Python Required Train models without writing code Unified Platform No jumping between tools Cloud GPUs Included No local GPU setup needed Instant Deployment Use your model immediately
State-of-the-Art AI Models

From Upload to Publication in 3 Steps

Designed for speed. Built for reproducibility.

1

Upload Your Data

Drag and drop microscopy files directly into your browser. Files are automatically converted for efficient streaming and metadata is extracted. Terabyte-scale datasets ready in minutes.

2

Create Workflow

Connect visual nodes—denoising, segmentation, tracking, quantification—without writing code. Each step runs on cloud GPUs. Iterate rapidly with instant previews.

3

Export & Share Results

Download publication-ready figures, CSV data, and annotated videos. Share workspaces with collaborators. Every analysis is versioned and reproducible.

From Images to Insights in Minutes, Not Months

Automated analysis pipelines that scale with your research. From developmental biology to drug screening, CellWorkflow accelerates discovery.

50×
Faster
10k+
Cells

Cell Migration & Tracking

Track thousands of cells across extended time-lapses. CellWorkflow generates complete lineage trees in under an hour—replacing weeks of manual annotation.

4TB+
Datasets

Light-Sheet & Organoids

Explore terabyte-scale light-sheet datasets directly in the browser with zero lag. 3D visualization enables real-time exploration of complex developmental processes.

384
Wells
90%
Time Saved

High-Content Screening

Standardize drug screening analysis across multi-well plates. Reproducible workflows ensure consistent results across experiments and research groups.

Why CellWorkflow?

Beyond the technology—real benefits for your research.

No hardware limitations

Process terabyte-scale datasets without upgrading your workstation.

No software licenses

Browser-based access eliminates expensive per-seat licensing.

Centralized data management

All datasets, workflows, and results in one accessible location.

Unified analysis pipeline

No more switching between fragmented tools and manual file transfers.

10× faster analysis

From weeks of manual work to hours of automated processing.

Reproducible results

Every analysis is versioned and shareable for transparent science.

State-of-the-Art Tools for Biological Research

Integrating cutting-edge AI models and scalable cloud infrastructure for reproducible analysis.

Z T Z-STACK & TIME-LAPSE

3D/4D Visualization

Render terabyte-scale datasets directly in your browser. Navigate through Z-stacks and time-lapses in real-time.

AI-Powered Segmentation

Deep learning models trained on millions of cells for accurate segmentation. Fine-tune on your own data for domain-specific results.

LINEAGE TRACKING

Advanced Cell Tracking

Automatic lineage reconstruction that handles cell divisions, merges, and complex behaviors across thousands of timepoints.

LOAD DENOISE SEGMENT TRACK VISUAL PIPELINES

No-Code Workflow Builder

Drag-and-drop nodes to create complex analysis pipelines. Save and share workflows across your team with one click.

DATA EXPORT

Quantification & Statistics

Extract morphology, intensity, and tracking metrics. Export to CSV or visualize with built-in interactive charts.

GPU GPU GPU AUTO-SCALING

Scalable Cloud Compute

GPU instances scale on-demand to match your workload. Process datasets far beyond your local hardware capacity.

NOISY CLEAN

Denoising & Registration

Deep learning denoising and registration algorithms clean up noisy images and align multi-channel stacks automatically.

A B C SYNCED

Real-Time Collaboration

Share workspaces with your team. See changes instantly with live synchronization. Comment, annotate, and analyze together.

.TIFF .CZI .ND2 ZARR STREAM TO BROWSER

Universal Format Support

Import TIFF, OME-TIFF, ND2, CZI, LIF, and more. Automatic format conversion enables efficient streaming to your browser.

Built by Researchers, For Researchers

CellWorkflow is developed by the imAIgene-lab at Instituto de Investigaciones Biomédicas (IIB), Universidad Autónoma de Madrid.

imAIgene-lab

Our lab leverages artificial intelligence and multi-omics analysis of dynamic 3D imaging to investigate tumor progression and treatment resistance. We bridge computational and experimental science to advance biomedical research.

Visit Lab Website →

Scientific Heritage

CellWorkflow builds on our proven track record in microscopy image analysis:

  • BEHAV3D – Published in Nature Biotechnology
  • BEHAV3D – 3D live imaging for T cell and tumor analysis (Nature Protocols)
  • BEHAV3D Tumor Profiler – Mapping cancer cell behavior in tumor microenvironment (eLife)

Principal Investigator

Dr. Maria Alieva leads the imAIgene-lab with expertise in AI-driven microscopy analysis for cancer research. Her work focuses on understanding T cell immunotherapy mechanisms and tumor microenvironment characterization.

IIB / Universidad Autónoma de Madrid

Research Roadmap

Planned capabilities and future research directions for CellWorkflow.

Foundation Models

Integration of next-generation foundation models for microscopy, including SAM 3 and domain-specific models trained on large-scale microscopy datasets.

Federated Learning

Privacy-preserving collaborative model training across institutions. Train better models without sharing sensitive data.

Spatial Transcriptomics

Integration with spatial transcriptomics platforms for combined morphological and molecular analysis.

FAIR Data Principles

Full compliance with FAIR (Findable, Accessible, Interoperable, Reusable) data principles for reproducible research.

Ready to Transform Your Research?

Start analyzing your microscopy data today. We're here to help you get the most out of CellWorkflow—reach out if you have questions or want to explore a collaboration.