Figure 3.

Fidelity of aligned tissue and identified neurons. (A) Representative example of sequential sections of the mouse brain identified using MCT. (B) Correlation of each section to the sections in front and behind with varying degrees of precision (as controlled by edge thickness) shows a tight alignment of individual coronal sections in the reconstruction. (C) Representative examples of neurons manually identified (red) and using the MCT algorithm (blue) near the injection site (top) and in a section with retrogradely labeled neurons. (D) Correlation of neurons identified by the MCT algorithm to the number of neurons identified by an expert human shows a high agreement (R² = 0.84). (E) Error of MCT algorithm in identifying neurons as a function of neurons labeled in the section shows that the MCT. consistently underestimates the number of labeled cells when the label density is high, likely due to poor segmentation of neurons near the injection site and when cell clusters are densely packed.