The analysis is tentative and not accurate based on several limitation.

1. Data

(1) Variety

Each case has variation. Some variation is minor, which will not affect Ordering or tracing. But in some cases, for example, there are so few vessels that no 6+ Order vessels can be found, causing large difference in result.

(2) Size

There are only 162 data ranging from 57 to 85, that is 5.6 cases in average in each age. If there is one strange data affecting the age group, the deviation will still be significant for lack of enough data.

(3) No continual observation for each individual

Analyzing each case in a certain age, then combining all the data to compare with each other is not an ideal method.to summarize brain changes through aging. The ideal method is collecting continual data from each individual through aging. However, such work is much heavier and difficult to realize. With the only data set, only a trend can be found in a rough idea.

(4) Human interference

It is hard to completely automatically trace all vessels in a brain. In BAFET, least human interference is required to get most standard and comparable results. But it is still based on a certain subjective aspect, such as deciding whether two vessels are linked together.

2. Calculation

(1) Linear model

It is assumed that all the results are based on linear relation when analyzing features, while its correctness remains unknown.

(2) Average

Features from each case are taken as a same weight in calculating average. Some bad or unrepresentitive cases are involved equally in the result.