A new method to monitor bone geometry changes at different spatial scales in the longitudinal in vivo μCT studies of mice bones
Fig 3
Evaluation of the algorithmic accuracy.
(A) Multi-resolution analysis of the Scenario 1: zero low- and hight-spatial frequency activity. (B) Analysis of the Scenario 2: only imposed low-spatial frequency activity. (C) Analysis of the Scenario 3: only imposed high-spatial frequency activity. (D) Analysis of the Scenario 4: imposed low- and high-spatial frequency activity. In each figure, from left to right, they show the decomposed coarse scale of geometric changes at different scenarios, the corresponding visualisation, the fine scale patterns and the corresponding visualisation. (E) The box plot of error distributions at different scenarios. The red line in the middle indicates the median value, the top and bottom of the box are the 75% and 25% of the error distribution and the whiskers extend to the minimum and maximum values. In sub-figures (A-D), the colourmap is used to indicate the magnitude and direction of the geometric changes. The warm colour (positive value) corresponds to the outward-pointing normal direction of the bone surface while the cold colour (negative value) corresponds to the inward-pointing normal direction.