Skip to main content
. 2019 Jul 13;9(18):5396–5411. doi: 10.7150/thno.28717

Figure 1.

Figure 1

Quantification algorithms. A: The structural algorithm performed color-based segmentation with k-means clustering on the red channel filtering out weaker intensities. The Hough transform detected the cells as bright circular objects on a dark background, providing their total count and the corresponding coordinates. Singular value decomposition (SVD) was used to filter the image followed by morphological operators that skeletonized and revealed the spine of the process. The coordinates of each cell were then used to initiate a stemming process search until an endpoint was reached. The distance of the endpoint to the cell center was measured. B: The intensity algorithm utilized the composite image as input to generate a hippocampal mask, omitting the irrelevant neighboring structures. The mask was applied to all three channels separately followed by color-based segmentation via k means clustering. For every channel the density is reported as the ratio of the pixels belonging to the cluster with the highest value over the pixels constituting the entire hippocampal mask. Scale bar, 100μm.