Intelligent Image Triggered Acquisition #
Abstract #
In bioreactor environments, the acquisition of live sample data often results in the collection of numerous frames with redundant or minimal information, leading to excessive time and energy spent on manual data curation by field experts. Current research in intelligent microscopy has primarily focused on improving the accuracy and speed of image analysis but has paid less attention to the efficiency and reliability of data acquisition itself. To address this gap, we develop the Intelligent Image Trigger Acquisition (IITA) system. This system employs an AI-driven decision-making framework to streamline the acquisition process. By incorporating an Efficient U-Net B5 architecture pretrained on ImageNet, the IITA system evaluates each frame in real-time to determine its relevance based on specific biological criteria such as cell count and viability. This approach significantly reduces the volume of non-informative data collected, thereby minimizing storage and computational demands. The system’s ability to selectively capture and store only pertinent frames reduces the need for extensive manual data curation, optimizing both time and energy resources. Results from live cell cultures in bioreactor settings demonstrate the system’s effectiveness in enhancing data quality and analysis efficiency. Future work will focus on expanding the system’s applicability to a wider range of cell types and biological conditions, further reducing the burden of manual intervention and paving the way for more autonomous biological research environments.
Presentation #
