Musculoskeletal Geometry

A 6-camera optical motion capture system (Oqus 5 +, residual error < 0.3 mm, Qualisys, Sweden) was used to track the positions of eight clusters, each comprised of three markers (Fig. 1), and a digitisation wand, comprised of five markers. Owing to the technical challenges of simultaneously tracking all 22 segments, the clusters were designed to be mounted and dismounted on different segments. An M4 ball and spring plunger system ensured accurate and repeatable cluster placement. A digitisation wand was used to digitise anatomical points, lines and surfaces on the specimen (error < 0.45 mm, as determined during the wand calibration).

Forearm specimen equipped with all the 22 cluster bases. The eight clusters are mounted on the humerus, the ulna, the radius, the second and third metacarpals, and the phalanges of the middle finger.

Initial dissection included the excision of skin, fat and connective tissues to reveal the bone surface to which the cluster bases were affixed. The bases were secured to each segment with self-tapping wood screws and care was taken to not impinge any tendons or ligaments. Six static trials were initially recorded, one for the forearm and one for each finger, to define a reference pose where the position of all clusters was known. The third metacarpal (MC3) cluster was used as the reference coordinate system. The error associated with the transformations of the marker clusters to the reference posture24 was less than 0.2 mm. All geometrical measurements presented correspond to this reference pose and are expressed in the anatomical coordinate system of the left ulna.29

Bony landmarks were digitized directly on the skin of the forearm and hand to define coordinate systems for all segments.29 The joints were passively moved through uniaxial motions to estimate the location and orientation of functional axes using the instantaneous helical axis method.26 The specimen was then dissected progressively to enable the digitisation of the origin, insertion and via points, representing the path of each muscle/tendon. Via points were digitised to describe more precisely the paths of the tendons between their origin and insertion points. These points were especially chosen to represent the paths of tendons when crossing a joint, e.g., entrance and exit to retinaculum, and to characterize the interconnection between finger tendons, e.g., extensors. After the dissection, the wand was used to digitise dorsal surfaces of the finger joints about which the tendons wrapped. Simple geometrical shapes, i.e., cylinders or spheres, were fit to these traces to further constrain the tendon paths in the resultant musculoskeletal model.