Table 1 The average expected error of a 3D parameter given a related 2D parameter estimated from a randomly chosen slice of a randomly chosen mitochondrion

From: A semi-automatic method for extracting mitochondrial cristae characteristics from 3D focused ion beam scanning electron microscopy data

Parameter

n = 1

n = 25

n = 50

n = 100

Mitochondrial Volume

85.01%

22.15%

15.71%

11.27%

Mitochondrial Surface Area

83.28%

19.49%

13.82%

9.97%

Intracristal Volume

86.10%

23.20%

17.13%

12.06%

Cristae Surface Area

88.12%

23.01%

15.92%

11.49%

  1. Rows give the Mitochondrial and Intracristal Volume in 3D as compared to corresponding 2D sectional areas, and Mitochondrial and Cristal Surface Area in 3D as compared to corresponding 2D sectional circumference. Columns give the number of randomly selected mitochondria used for each estimate. E.g., when estimating the functional relation between the true Mitochondrial Volume V and the corresponding 2D area A from 25 randomly chosen mitochondria, we assume a conditional normal distribution P(V∣A) = N( μV (A), σV (A)) and estimate \(\frac{{\sigma }_{V}(A)}{{\mu }_{V}(A)}=22.15 \%\). This means that an estimation of the average mitochondrial volume based on measurement of 2D cross-sectional area on a single random slice through each of 25 random mitochondria can be expected to be 22.15% erroneous. Please refer to 2D to 3D relationships for details and in particular Supplementary Fig. 3 which shows the corresponding μ and σ curves.
Back to article page