Quantitative analysis and classification of gait patterns in cerebral palsy using a three-dimensional motion analyzer

Gait disorders in cerebral palsy can be accurately analyzed using the CODA-3 system presenting quantitative data representing movement of the hip, knee, and ankle in the sagittal plane. We describe a technique that classifies abnormal gait automatically on the basis of sagittal kinematic data. Fifty-five hemiplegic and 91 diplegic patients were analyzed using an opto-electronic scanner (CODA-3). The sagittal kinematics of the affected limb in hemiplegics correlated with those of both affected limbs in diplegics. We introduce the concept of the 'plegic limb.' Sagittal kinematics of 237 affected limbs were studied using cluster statistical analysis. Eight clear groups emerged. The predominant clinical features, typical of each group, were identified and described (eg, stiff leg gait, genu recurvatum, or crouch gait). We propose this classification system as a new technique to use gait analysis data to automatically classify abnormal movements of the lower limb in cerebral palsy.