Laboratories employ various techniques to assess the readiness for passage, including microscopic observation and cell counting. While microscopic examination helps evaluate cell morphology and confluence, it is a monotonous, manual work that can be very subjective and prone to error without any external guidance or experience. There is a need for a fully automatic tool in order to eliminate the risk of late passage, automate and speed up the process, as well as mitigate subjectiveness of observations.
CellONE is a cloud based an algorithm that automatically analyzes the image obtained with a microscope utilizing Artificial Intelligence, Convolutional Neural Networks and Computer Vision in order to obtain a high-quality cell detection and analytics. Based on the image context, like cell confluence percentage, cell clustering detection and cell merging, the algorithm is able to recommend further actions for optimal cell culture growth. Results can be visualized as a segmentation mask with cells as foreground, precise localization of major cell clusters with a bounding box, or a set of parameters in the form of a pdf report.
The analysis can be executed by simply dropping the image on the website, or by including it locally in the lab in an automatic data processing pipeline, where human observations are only optional. At its base it can increase reproducibility, reliability and save time and improve laboratory capacity by not requiring an expert to manually observe and make decisions for each cell culture sample. The system can keep track of cell culture growth and alert the lab when a specific sample needs their attention. Furthermore, retrospective data from the lab can be used to fine-tune the decision-making process for the algorithm to mimic specific required culture growth standards. With the addition of a robotic arm and camera system, the whole process of passage and cell culture growth in general can be fully automated and sped up.