Cellular spheroids: How to rely on them by non-destructive methodologies
The use of 3D cellular models have been developing very fast, among the methodologies for biofabrication of this models, most of them have the final application as components of new materials – cell interaction or drug screening. The spheroid shape is the easiest form to have a simple 3D model. However, this model lacks standardization of its characteristics for further use in biofabrication. Because of this, the aim of this work is to give a set of index that can be used as non-destructive way to keep the spheroid for other applications. The methodology chosen for spheroids preparation was the non-adhesive using micromolds with the cell line NHI-3T3. It was varied the initial number of cells to make the spheroid, they were imaged in 3 and 7 days of culture. The spheroids in the images were analyzed according to the ASTM F1877-16 were the index Equivalent Circle diameter (ECD), The Aspect Ratio (AR), Roundness (R) and Form Factor (FF) were calculated. Convolutional Neural Networks (CNN) were also applied for identify live and dead cells. As results, the spheroids formed with the initial amount 100,000 cells presented the best indexes to spheroid formation, more than this the spheroids had an elliptic form and less than this is observed a cluster formation at 7 days of cultivation. Among the CNNs used to classify the dead and live cells in the images, the AlexNet was the most suitable for this application. This initial characterization applying the ASTM-F1877-16 showed to be suitable for initial spheroids parameters, which allow qualitative data begun quantitative. When spheroids are used for this application, as for example in bioprinting of translucent materials, followed by AlexNet CNN, it is possible to have a full new material analyzed and characterized prior to the use without losing any sample.