Related Experiment Videos
Classification of resistance to passive motion using minimum probability of error criterion
H C Chan1, M T Manry, G V Kondraske
1Department of Electrical Engineering, University of Texas at Arlington 76019.
Annals of Biomedical Engineering
|January 1, 1987
Summary
A new computerized technique accurately classifies normal and Parkinson disease patients by analyzing knee joint resistance to passive motion. This automated method shows over 95% agreement with physician assessments, aiding in neurological disorder diagnosis.
Area of Science:
- Neurology
- Biomedical Engineering
- Medical Diagnostics
Background:
- Neurologists assess limb resistance to passive motion for diagnosing muscular and nerve disorders.
- Automated measurement and parameterization of limb resistance have been developed using computer-based instruments.
- A motorized driver moves a relaxed lower extremity at constant velocity, sampling torque and angle.
Purpose of the Study:
- To describe a computerized technique for classifying patients as 'Normal' or having 'Parkinson disease' (rigidity).
- To analyze the torque versus angle curve of the knee joint for diagnostic classification.
- To compare automated classification results with independent physician assessments.
Main Methods:
- A computer-based instrument measures torque and angle of passive lower extremity motion.
- A Legendre polynomial is fitted to the torque-angle curve.
- Eight normally distributed features are calculated from the polynomial fit.
- A minimum probability of error approach is used for classification.
Main Results:
- The computerized technique successfully classified patients based on knee joint torque-angle curves.
- Data from 44 subjects were processed.
- The automated classification achieved over 95% agreement with subjective physician assessments of rigidity when all features were used.
Conclusions:
- The described computerized technique provides an accurate and objective method for diagnosing Parkinson disease-related rigidity.
- This automated approach has high concordance with clinical judgment, potentially improving diagnostic efficiency.
- The method offers a reliable tool for the quantitative assessment of neuromuscular disorders.