Whiplash-associated Disorders

Abstract

Introduction Whiplash-associated disorders (WAD), as defined by the Quebec Task Force, are a common diagnosis after neck trauma, caused by sudden acceleration and deceleration forces acting on the head and neck, most typically related to rear-end or side-impact car accidents. Since physical injuries seldom are found with present imaging techniques, the diagnosis can be difficult to make. The active range of motion of the neck is often visually inspected in patients with neck pain, but this is a subjective measure, and a more objective decision support system, that gives a reliable and more detailed analysis of neck movement pattern, is needed. The objective of this study was to evaluate the grading ability of a unsupervised self organizing maps (SOM) network when presented with neck movement patterns as input and using the individual subjective Neck pain and disability index (NPAD) as an indication of the condition of the subject. Method Three-dimensional neck movement data from 59 subjects with WAD and 56 control subjects were collected with a ProReflex system. Rotation angle and angle velocity were calculated using the instantaneous helical axis method and motion variables were extracted. A principal component analysis was performed in order to reduce data and improve the SOM network?s performance. Each subject had to fill in a questionnaire, before the study begun, which was used for the calculation of the NPAD index. One vector, containing the movement pattern, for each subject was then presented to a SOM network, configured with different number of hidden nodes, during a training session. The NPAD index was used in combination with the classification of the SOM network in order to visualize the result. Each hidden node represents a class with in the SOM network e.g., if a SOM network has two hidden nodes it tries to classify the movement patterns into two categories. Results and conclusion The results show that when using a SOM network with two hidden nodes it is possible to classify the subjects as control or WAD with a predictivity of approximately 85 %. If only WAD-subjects are used as input to a SOM network with two hidden nodes there is a classification of subjects with NPAD greater 60 in one category and those with less than 60 in the other category. In conclusion, this method seems promising when developing a tool for grading of WAD but further evaluation is needed.