To end this tutorial, here is a discussion of the findings shown on these graphs.

There is a gradual increase in particle (red blood cell) size in the majority of the middle part of figure 9. In addition, there is a steep increase in area at both the beginning and end of the graph (fig. 9). Since the particles were arranged in ascending order based on area, this suggests that particle area may follow a normal distribution. Figure 10 was created to test this, and as illustrated, a highly symmetrical normal distribution curve is created when an area frequency graph is generated. There are 2 particles in the 40-60 um^2 range (fig. 10) which results in a slightly negative skew in the graph. These two particles could however be outliers within the analysis that are not red blood cells but smaller fragments of red blood cells or background noise within the image that should have been excluded in the analysis. This, perhaps, shows one the key limitations of imagej, where any analysis is only as reliable as the image taken into the analysis. If, for example, the thresholding process, or image restriction process was not done adequately, a significant more amount of stray particles would have been included into the raw results. Even in this example, it is hard to determine whether these outliers are actually red blood cells or unrelated particle fragments. To continue, figure 11 shows that red blood cells are most clustered in a specific area and circularity value. As this specifc area and circularity value increases and decreases, the concentration of particles in the graph gradually declines in all directions. The position of the outliers in relation to all the other particles is shown as well; the relatively low value of circularity of outliers in the 40-60 um^2 range further suggests that they are possibly unrelated particle fragments instead of red blood cells.