Tips and Tools : Precision testing using frozen cadavers

The precision of an XROMM study can be quantified by substituting a frozen specimen for the live subject in the workflow. This method is study-specific and should be repeated for each XROMM study conducted.

A frozen specimen is mounted to a substrate that is not radiopaque (e.g. a wooden pole) and then moved in three dimensions within the biplanar x-ray field. Joints in a frozen specimen should be immovable and thus by default all relative motions between markers and bones, respectively, should be zero. The amount of deviation from zero is the measurement of precision.

The current method quantifies precision using standard deviations of the resulting intermarker distance and the 6 DOF data (translations and rotations of joints). Movements recorded in live subjects that fall below these standard deviations should be considered error or “noise” and not interpreted as real joint motions. The following examples demonstrate precision measurements in various XROMM studies.

Example data set

Frozen Pig Head

The frozen pig specimen had a total of 9 markers: 1-4 in the skull and 5-9 in the mandible. Markers from 2 trials (1,000 frames/trial) were digitized and standard deviations of intermarker distances calculated.

Rigid body kinematics were calculated from digitized marker XYZ coordinates and used to animate bone movements. A joint coordinate system (JCS) was set for the pig mandible using the mandibular condyles to align the JCS. 6 DOF data was generated from this JCS and standard deviations were calculated for each aspect of joint translation and rotation.

The results of this precision study can then be applied to 6 DOF data in order to distinguish noise from real movements. Here, error envelopes have been added to a chart of mandibular rigid body translations during pig chewing. The error envelopes are calculated as the mean of the translational movements (for this frame subset) +/- the SD for that translation as calculated in the precision study. For example: the mean of T(X) in these frames +/- precision study SD for T(X) = 0.190 +/- 0.005 cm.