Version 1.6 by puchades on 2020/03/25 14:47
Show last authors
1 == Analysis approach for series of rodent brain section image ==
2
3 There are two main approaches for the analysis of rodent brain section images.
4
5 1. Pixel classification only (with two or more classes)
6 1. Pixel classification with two classes (//immunoreactivity// and //background//), followed by object classification with two classes (//objects-of-interest// and //artefact//).
7
8 **Which approach is best for my dataset?**
9
10 As a general rule, pixel classification is suitable for images in which there are clear differences in the colour, intensity and/ or texture of the feature-of-interest versus the background and other structures.  If there is non-specific labelling in the image that is very similar in appearance to the labelling-of-interest, object classification may allow the non-specific labelling to be filtered out based on object level features such as size and shape. The best approach is determined by trial and error.
11
12 === Pixel classification workflow ===
13
14
15 ==== ====
16
17
18 Photo by David Clode
19
20 ==== Or code ====
21
22 Code blocks can be added by using the code macro:
23
24 {{code language="python"}}
25 x = 1
26 if x == 1:
27 # indented four spaces
28 print("x is 1.")
29 {{/code}}
30
31 (% class="wikigeneratedid" id="HH4Won27tAppearinToC" %)
32