br Results br Discussion As previously reported

Results
Discussion As previously reported, human based hNSC fate analysis was found to be a very subjective and inconsistent method even when human observers are experienced. The greatest variation in human based manual cell quantification both between individuals and between repeated analyses by a single individual was found in the number of bromodomain positive only for Hoechst, GFAP, and ß-TubIII and GFAP double positive immature progenitors, suggesting that humans have a limited capacity to apply an accurate reference baseline to object volumes and different color intensities, e.g. for the size of nuclei and different levels of red and green mixed together. Low internal and external validity of human based image analysis can affect interpretability of human stem cell in vitro differentiation data. Conversely, Volocity® software made some quantification errors, especially when two cytoplasmic cell fate markers together with nuclear counter staining were analyzed at once. The software based quantification errors were consistent and the number of the errors was cell density dependent. A human operator “data validation” step to confirm Volocity® object recognition and correct mislabeled objects manually gave higher ≥94.4% accuracy, but also increased the time of the analysis. Nonetheless, Volocity® software based, operator validated, hNSC fate classification and quantification was still 2 fold faster than the human based quantification alone. Plating the cells in lower density may also overcome the problem with the mislabeled objects. However, when working with certain types of cells, such as hNSCs, this can be technically challenging because of the relatively long in vitro differentiation period required or the practical limitation that the cells prefer growing in high density. Protocol modifications, such as adding “erode objects” decreased the number of software based quantification errors, but at the cost of losing some correctly labeled cells. The software based object separation was found to be less demanding if using only one cell fate marker at once or if the objects of interest were strictly overlapping each other, e.g. nuclear Olig2 and Hoechst, which in some cases allows software based image analysis in 2D. The validation of software based object recognition is highly recommended for better accuracy, especially when quantifying cells in high densities. Finally, Volocity® based cell image analysis requires careful standardization of immunocytochemistry protocols and image acquisition parameters in order to avoid unnecessary data variations. Despite of these limited disadvantages, when high numbers of samples need to be quantified, Volocity® based cell quantification can improve both productivity and interpretability of in vitro differentiation data across different researchers and research groups.
Materials and Methods
Conclusions Accurate and efficient quantification of human stem cell lineage and fate is critical for the many assays required for the translation of basic stem cell research to clinical therapeutics for disease or trauma. Conventional human based manual cell fate analysis is not just time consuming but a subjective and inconsistent method. The Volocity® software based human neural stem cell classification and quantification protocols presented in this paper allow for semi-automated, 2 fold faster, and more accurate image analysis of cytoplasmic and nuclear cell fate markers after in vitro differentiation. These factors improve both productivity and interpretability of in vitro differentiation data across different researchers and research groups. Altogether, these data suggest that Volocity® image analysis software can be used as a precise tool in conjunction with both inverted and confocal laser scanning microscope image acquisition platforms. The Volocity® software protocols presented in this paper are not limited only to human neural stem cell fate analysis or the image acquisition platforms used in this study; with small modifications, such as adjustments in object size, color intensities, and z-step interval, these protocols can be used for quantification of other humans stem cell lineages and in conjunction with other microscope image capture systems.