We were encouraged in our modeling approach by recent observations we made in patients with Parkinson's disease. The cardinal symptoms of these patients are hypokinesia and rigidity, but also stance control is impaired. The impairment includes abnormally large and abnormally fast spontaneous body sways. This appears to be paradoxical, because the rigidity entails an increased axial stiffness, which should lead to smaller rather than larger sways. A second apparent paradox is found after treatment (here with dopamine medication and/or subthalamic nucleus stimulation). This leads subjectively, for the patients, as well as on clinical rating scales to an improvement, yet sway amplitude remains abnormally large or even becomes larger (see Maurer et al., 2003).
In a more detailed analysis of these findings (Maurer et al., 2004), the patient's sway revealed an abnormally strong tendency for body oscillations at about 1 Hz. Simulations of our model allowed us to mimic this abnormal sway behavior. This applied when we enhanced the loop gain and assumed an abnormally large internal noise in the system. Thus, a form of resonance in the sensory feedback systems was able to explain the abnormally large and fast sway of the patients (noise components in the order of the 'eigen frequency' of the system produce the resonance tendency). Furthermore, the paradoxical therapy effect could be mimicked. This applied when the loop gain in the model was reduced to normal levels, while the abnormally large noise remained. Even idiosyncrasies of the individual patients' therapy effects could be mimicked.
Meanwhile we have established a method of parameter identification for non-linear models (non-linear because of the aforementioned thresholds in Fig. 16-5) and have successfully applied it to experimental data (Maurer et al., 2006). We hope to use this method now to identify the source of the abnormally large noise in the patients' stance control. From preliminary data it appears that a main source is the vestibular signal and that the ability to replace it by other signals through sensory re-weighting is impaired in the Parkinson's disease patients.
Was this article helpful?