AI and Automation

Research in this topic focusses on the development of new methods and technologies for cellular disease models. This includes: the development of AI-based microscopy image analysis, lab automation for screening of 2D/3D cell models and organ-on-chip models, and the full integration of these techniques for autonomous experimental set-ups (smart screening).

Artificial Intelligence for Microscopy

Microscopy plays an important role in the phenotypic discovery of cellular disease models. Standardized and trustworthy analysis of the microscopy output is essential for meaningful phenotypic discovery. This becomes even more important for large (screening) data sets when manual analysis would be too cumbersome and labor-intensive. We develop various Artificial Intelligence based tools and analysis pipelines to overcome these issues and allow for advanced phenotypic discoveries.

AI focus

Our work includes (but not limited to) AI based:

  • 2D and 3D individual cell or organoid segmentation
  • 2D and 3D shape analysis
  • unsupervised or supervised feature and phenotypic discovery
  • label-free imaging
  • image restoration.
Our work on Github

Lab automation

Cutting-edge pre-clinical (human) cellular disease models play a critical role in understanding biology and for the development of novel therapeutic strategies in other disease areas. However, for the translation of these on the bench disease models to trustworthy models that impact patient care additional steps of standardization, automated screening, and increased throughput are required. We develop automated methods and technologies for the standardization and upscaling of cellular disease models.

Lab automation focus

Our focus includes (but not limited to):

  • Development of automated human cellular disease models (from bench to automation)
  • Microscopy based medium- to high-throughput cellular screening (2D/3D, high-content, live-cell)
  • Compound and getentic preturbation studies
  • Development of multi-organ-on-chip technologies
  • Consultancy in pre-clinical lab automation including experimental design

We facilitate researchers on Utrecht campus on lab automation within the Cellular Screening Technologies facility.

Smart Screening

Using the robotics framework of “Sense-Think-Act” we integrate on-the-fly analysis with feedback microscopy, screening, human oversight and validation. This allows for the execution of full autonomous experimental set-ups with human oversight: from a tube with cells to an validated and trustworthy analysis report. Our research focuses mainly on the development of methods and technologies that allow for these autonomous pipelines using modern IT and AI tooling.  We work closely together with SURF in using public cloud solutions and tools.

Cellular Disease Models Research Group Team

Collaborations

We are open to help academic and biotech researchers in their questions and need for advanced methods and technologies. Feel free to contact us.

Key publications AI & Automation

Active mesh and neural network pipeline for cell aggregate segmentation. Smith MB, Sparks H, Almagro J, et al. Biophys J. 2023;122(9):1586-1599.
OrgaSegment: deep-learning based organoid segmentation to quantify CFTR dependent fluid secretion. Lefferts JW, Kroes S, Smith MB, et al. Commun Biol. 2024;7(1):319.

AI and Automation Team

Our experts

Sam-van-Beuningen
Dr. Sam van Beuningen

Assistant professor & Principal Investigator

s.vanbeuningen@umcutrecht.nl Research profile
Dr. Matthew Smith

Senior postdoc computational image science

Research profile
Dr. Tijmen Booij

Senior postdoc lab automation

Roos-Anne Samsom portrait
Roos-Anne Samsom

Senior research technician lab automation

Research profile
Lianne Winkel portrait
Lianne Winkel

Senior research technician lab automation

Research profile
Bram Bosch portrait
Bram Bosch

PhD student AI for microscopy image analysis

Research profile
Krijn van der Steen portrait
Krijn van der Steen

PhD student Smart Screening and Microscopy

Research profile