Cell culture monitoring is an essential clinical process in which cultivated cells are studied and monitored for deriving key conclusions about health and wellbeing. The legacy approach to the process involves significant human intervention, which can lead to cell damage and resultant errors in the results derived.
In this case, continuous cell monitoring over a sequence of time is not possible and is very challenging for the clinical operators to collect the number of cells samples and cultivate them in a safe environment.
The present paper proposes a solution to provide an efficient non-invasive cell colony observation application. It includes cell colony segmentation and its behavior (i.e., a healthy or unhealthy cell colony) recognition through a set of time series videos. The solution will also increase the operator performance in cell quality analysis and help minimize the cell wastage by combining computer vision and AI-driven pattern recognition algorithms to speed up operational time.