Fig. 3
From: Sensitivity of CNN image analysis to multifaceted measurements of neurite growth
NeuriteNet scoring by pattern is partially explained by neurite tracing data and clusters with Alignment Index. A Scatter plot of replated DRGN plotted by their Predicted Pattern Score and rank normalized Alignment Index. Of note, for these approaches all DRGNs with tracing data were used. The scatter plot shows strong clustering between the Predicted Pattern Score and Alignment Index with Score increasing with Index decreasing. B Comparison of how well the two-machine learning approaches are explained by the tracing data via a ridge regression model of the tracing data seeking to explain the variance of the Predicted Pattern Score. This shows that NeuriteNet classification scoring is partially explained by the tracing data