Abstract
Aims: To detect discriminant features in the cyclic kinematic patterns generated during selected upper limb activities of daily living by a normative and a distal radial fracture group, so as to reduce the multidimensionality of the kinematic analysis.
Methods: Cyclic activities of daily living were performed using a protocol that allowed comparison between the resulting kinematic patterns or waveforms. Twop groups were measured:
Group A: 11 subjects with normal hand function (average age: 31.5ys, SD: 8.7ys).
Group B: 5 subjects having undergone treatment for distal radial fracture was tested using the same methods (average age: 34.2ys, SD: 16.8ys).
Task presented here, (one of 5) performed by turning a key 90 degrees clockwise. Principal component analysis (PCA) was applied to the waveforms of group A, using the procedure illustrated by Deluzio et al.,1997 for use with walking gait patterns. A 90% trace criterion was used to calculate the number of principal components (PCs) to retain.
Results: Looking at elbow pronation/supination (PS). Two PCs were retained. The first component consisted of a simple pronation pattern. The opposite signs of Y1 differentiated left-hand users (utmost right), who required pronation to rotate the key, from right-hand ones, who required supination, with the exception of subject 3 group B. The second component consisted of pronation (cycle first half) followed by supination (second half). Subject 3 stood out because of limited elbow supination, which resulted from the combination of pronation (Y1) and supination (Y2) components.
Conclusions: Upper limb analysis can employ the statistic tools of gait analysis provided a cyclic and repeatable protocol is used. PCA was applied to elbow PS to identify statistically different movements of the distal radial fracture group and underline their main characteristics. This is particularly important in the presence of a large data group, when the identification and evaluation processes need to be both rapid and accurate. Limited PS was identified as a discriminant feature, supporting the follow-up studies for this injury that measured a reduction of PS by about 80% compared to that of the unaffected side. The cycle stages concerned can be identified on the basis of the contribution given by each component.
Correspondence should be addressed to Mr Carlos Wigderowitz, Senior Lecturer, University Department of Orthopaedic and Trauma Surgery, Ninewells Hospital and Medical School, Dundee DD1 9SY.
