Skip to main content

Table 1 Linear regression analysis results for hexagon size = 825 pixels (≈0.75 mm TMA core)

From: Impact of tissue sampling on accuracy of Ki67 immunohistochemistry evaluation in breast cancer

R2 values
HexN All tumor cases Homogeneous cases Heterogeneous cases
Sum Mean Median Sum Mean Median Sum Mean Median
1 0.827 0.827 0.827 0.906 0.906 0.906 0.6 0.6 0.6
2 0.893 0.888 0.888 0.929 0.926 0.926 0.749 0.733 0.733
3 0.93 0.92 0.9 0.964 0.955 0.947 0.81 0.79 0.731
4 0.955 0.952 0.945 0.969 0.969 0.965 0.897 0.879 0.866
5 0.964 0.958 0.938 0.972 0.969 0.963 0.926 0.912 0.842
6 0.964 0.957 0.951 0.975 0.972 0.969 0.916 0.895 0.886
7 0.964 0.959 0.952 0.976 0.969 0.969 0.918 0.913 0.883
8 0.969 0.958 0.95 0.981 0.977 0.977 0.916 0.882 0.845
9 0.976 0.971 0.966 0.984 0.98 0.981 0.944 0.934 0.908
10 0.978 0.971 0.962 0.987 0.983 0.98 0.947 0.923 0.893
11 0.974 0.968 0.961 0.983 0.978 0.977 0.936 0.92 0.894
12 0.977 0.971 0.968 0.982 0.976 0.977 0.954 0.941 0.926
13 0.982 0.978 0.969 0.986 0.984 0.984 0.965 0.95 0.914
14 0.98 0.975 0.969 0.986 0.984 0.984 0.952 0.938 0.907
15 0.982 0.977 0.973 0.984 0.983 0.983 0.965 0.95 0.933
  1. In each data set Ki-67 LI was calculated by counting mean, median and sum of positive and negative cells. All linear regression analysis results were statistically significant, p < 0.0001