In addition to providing image editing facilities, the PN program allows to do image quantitation. The principal motivation for quantitations in the design of the PN programs have been to examine how neurons are affected by the spatial location of the AB peptides in Alzheimer patients.
In the picture below a dialog presents two graphs, blue for the left picture and red for the second picture, that denotes the histogram of pixel values. For example, the blue curve shows that on the neuron picture (left, below) there is an increased number of pixels of value 50, as opposed to the red curve that indicates an almost monotonic fall of pixel population from value 0 to 255. This graph is useful in determining the quality of the image and helping in the implementation of the image editing filters.
The picture below has suffered several steps before getting to the final state. First, a digital image for the AB peptide was read in the (right) picture of the program, secondly a wavelet filter was applied in order to get rid of high-frequency noise. Then the thresholding filter was applied in order to separate the most intense regions. The PN program then offers the facility of identifying disconnected clusters and labeling them according to size and marking them with a green dot that represents the location of their center of mass.
The identified clusters in the picture can be (mouse) selected and marked (green) in order to edit them in the picture.
All or selected clusters in the pictures are analyzed and their properties can be listed in the data dialog. The clusters can be characterized by their radius (Rg), total mass, density and area. Hidden graphs in the same dialog can plot histograms of the mentioned properties.
Local analysis between the two pictures can be realized by separating the neurons in the left picture, for example, and measuring how mush of the right picture is locally surrounding them. The red rings indicated neurons (left) that are surrounded by a lot (above a certain value) of the AB peptide (right).
The rings in the figure above can be classified according to how much of the AB is found in them and a histogram can be drawn (below) to rank the number and how much the neurons are surrounded by the AB. The dialog below provides parameter adjustments to tune the threshold criteria, as well as the sizes of the rings around the clusters (neurons) in the left picture.
One can sit at the center of the AB clusters and map in 2 dimensions how the pixels decay in value. After doing one map per cluster and averaging over all of the cluster, an average 2 dimensional map results. The map indicates a dark center (left) that indicates that the pixel value decays rather rapidly from the center (right).