AI-powered Smart Microscopy for Stem-Cell Engineering

Cell engineering is predicted to play an important role in the future of regenerative medicine. Human mesenchymal stem cells (hMSCs) are a cell type of major interest and can be obtained from sources including bone marrow, adipose tissue and umbilical cord. More than 900 hMSC clinical trials have been conducted since 2004, alongside thousands of academic publications exploring their therapeutic potential. Despite this substantial research effort, relatively few hMSC-based therapeutic products have reached clinical use, highlighting the challenges of translating promising cell therapies into robust, reproducible and scalable treatments.

Our goal is to introduce a new paradigm in stem-cell engineering, replacing the traditional trial-and-error approach with quantitative, data-driven and continuous optimisation. We are developing an AI-powered smart microscopy platform in which advanced imaging is combined with machine learning to monitor how stem cells respond to different culture conditions and identify the cellular features associated with desirable outcomes.

This approach generates large volumes of complex microscopy data, creating new computational challenges and opportunities. Computational methods and high-performance computing are becoming increasingly important in modern microscopy and cell biology, enabling researchers to process large datasets, extract meaningful information from images and train increasingly sophisticated machine learning models. Rather than analysing images individually, AI models can be trained on thousands or millions of cellular images to identify subtle patterns in cell morphology, behaviour and function that may not be readily apparent to the human observer. Computational resources are therefore essential for developing, training and evaluating these models efficiently.

Ultimately, we aim to create a next-generation bio-manufacturing process in which cell stimulation can be dynamically adjusted according to the real-time behaviour of the cell population. By combining imaging, machine learning and automated sample manipulation, the system can analyse cells, make data-driven decisions and trigger experimental interventions with minimal human input.

The ability to combine different imaging technologies, trigger imaging modalities based on real-time image analysis, and synchronise image acquisition with external devices such as liquid-handling and cell-manipulation systems opens the possibility of fully automated and adaptive cell-engineering workflows. The integration of advanced microscopy with computational modelling and AI has the potential to transform how biological experiments are designed: moving from observing cells to continuously learning from their behaviour and using that information to optimise the next experimental step.

For more information about the project contact Dr Giuseppe Di Caprio (giuseppe.dicaprio@strath.ac.uk) Strathclyde Chancellor’s Fellow at the Department of Biomedical Engineering at the University of Strathclyde.

For a list of the research areas in which ARCHIE-WeSt users are active please click here.