Multiple meals are sequentially imaged and the data is uploaded to the server that performs computer vision processing, predicts when cells will exceed a pre-defined threshold for optimal cell confluency, and provides a Web-based interface for remote cell culture monitoring. the Metroprolol succinate degree of cell confluency with a precision of 0.7910.031 and recall of 0.5590.043. The system consists of an automated phase-contrast time-lapse microscope and a server. Multiple dishes are sequentially imaged and the data is uploaded to the server that performs computer vision processing, predicts when cells Rabbit Polyclonal to NBPF1/9/10/12/14/15/16/20 will exceed a pre-defined threshold for optimal cell confluency, and provides a Web-based interface for remote cell culture monitoring. Human operators are also notified via text messaging and e-mail 4 hours prior to reaching this threshold and immediately upon reaching this threshold. This system was successfully used to direct the expansion of a paradigm stem cell population, C2C12 cells. Computer-directed and human-directed control subcultures required 3 serial cultures to achieve the theoretical target cell yield of 50 million C2C12 cells and showed no difference for myogenic and osteogenic differentiation. This automated vision-based system has potential as a tool toward adaptive real-time control of subculturing, cell Metroprolol succinate culture optimization and quality assurance/quality control, and it could be integrated with current and developing robotic cell cultures systems to achieve total automation. == Introduction == The use of stem cells forin vitromodels of biological processes or forin vivocell-based therapies typically requires total initial cell numbers that exceed those normally available from a single isolate of main cells[1],[2],[3],[4],[5],[6]. To produce sufficient numbers of cells requires first inducing proliferationin vitroutilizing standard subculturing processes whereby cells undergoing proliferation in each culture vessel are periodically subdivided and re-plated into multiple vessels through several passages[1]. The decision on when to passage cells is currently based on a human operator’s visual assessment of cell confluency, which refers to the amount of space in a tissue culture vessel that is occupied by cells and reflects cell population density. Predetermined schedules of time-points for subculturing might be sufficient for growing well characterized, established cell lines[7],[8],[9],[10]. However, in general, unpredictable changes or disturbances in culture conditions[11]or large variations in isolate-to-isolate applications of main cells[12],[13]dictate that subculture be adaptively decided on-the-fly by direct observation of confluence over time[14]. Traditionally, human operators manually estimate confluence by microscopic observations and subsequently decide on the appropriate time for performing subculture. Presently, the majority of automated or semi-automated cell culture systems that are commercially available or in development still rely on either human oversight or a pre-determined routine to monitor cell cultures[7],[8],[9],[10],[14]. While there are systems that use Metroprolol succinate electrical impedance measurements of the cell-substrate as an indirect but automatic measure of confluence[15], some human oversight will still likely be required to monitor the process, including observing cell density and morphology to ensure optimal culture quality. The use of human operators to make decisions on subculturing is usually highly subjective and prone to intra- and inter-operator variability[7]. And, in the production of clinical-grade cells, the high cost of experienced labor substantially increases the costs of quality control (QC) and quality assurance (QA) operations[16]. Furthermore, it is not practical or cost-effective for human operators to manually observe and monitor cell cultures continously, and therefore key events such as the optimal times to perform subculture or identify problems might be missed. Delayed subculturing can result in cell overgrowth, which leads to loss of stem cell differentiative potential or stemness[11],[17], whereas premature subculturing can lead to longer production times to achieve targeted cell yields, with associated added costs. The overall lack of reproducibility and control of clinical-grade cell expansion processes is usually a major concern of authorities regulatory body since this has a direct impact on product performance and product reproducibility[18],[19],[20]. In addition, the lack of subculture standardization and reproducibility hampers scaled, robust and cost-effective manufacture of cells and has been cited as a major hurdle in the development of stem cell engineered products[7],[16]. Consequently, whether using a manual or robotic cell culture system, there is a need to automate monitoring Metroprolol succinate of and decision-making for the subculturing process[16]. To begin to address this need, machine vision technology has been applied to detect cells and measure Metroprolol succinate confluence to determine the appropriate time to culture cells[17],[21]; however, the images derived from this system are similar to that of a brightfield microscope and as such, of low-contrast[21], making it hard to verify cell detection overall performance. Additionally, this system did not incorporate real-time predictive modeling of cell growth, and lacks the capability to function as a part of a QA/QC system by raising warning alarms if growth was not progressing as expected and, in manually operated systems, as a tool to alert human operators in a timely manner to make preparations for subculture. Herein we statement on a new technology platform for continuous, fully automated monitoring, analysis, and predictive growth modeling of phase-contrast time-lapse microscopy imaging of the subculturing process. This platform is based on our previously developed real-time, computer vision-based cell tracking system[22],.