Simple assessment for rotation of the forearm using three-dimensional bone models

Background

Dynamic medical imaging can determine the cause of rotational impairment in the forearm. However, it has drawbacks depending on the image modality used, related to radiation dose, the need for specialized equipment, and the labor intensity involved in the analysis. Because the forearm rotation axis is static, we hypothesize that an axis based on bony landmarks is comparable to an axis calculated from dynamic imaging.

Methods

Eight post-mortem human forearms were scanned using CT in seven rotational positions from maximum supination to maximum pronation. Three rotation axes were calculated: the landmark, average helical, and circle fit axes. The primary outcome is the difference between the axes expressed as the angle and the minimal distance between them. Secondary outcomes are the orientation errors when modeling pose using the three found axes.

Findings

The mean difference between the landmark and average helical axes was 0.38 degrees and 0.51 mm. The mean difference between the landmark and circle fit axes was 0.40 degrees and 0.51 mm. When modeling the pose of the radius using one of the three axes, the difference between the modeled radius and the scanned radius was in each direction below 2 mm and 1 degree.

Interpretation

The rotation axis of the forearm can be accurately calculated using automatically placed bony landmarks. These findings indicate that determining the forearm rotation axis does not require multiple static images or dynamic imaging. This knowledge should be applied to clinical data to assess its applicability in practice.