**7. Conclusions**

In this paper, a self-managed system for the automated assessment of Parkinson's disease at home is presented. The automated assessment is focused on lower limbs, posture and postural stability tasks as specified by standard clinical assessment scales. A high usability of the system is guaranteed to motor impaired users by a gesture based human computer interface. The patient movements are characterized by sets of selected kinematic parameters which best correlate with clinical UPDRS scores, collected in an experimental campaign conducted on PD subjects. The data acquired have been used to train supervised classifiers employed for the automated assessment of new task instances. For the first time, in the context of Parkinson's disease, low-cost optical tracking devices are used to characterize center of mass movements as an index of postural instability. Preliminary results on the assessment accuracy, as compared to standard clinical evaluations, suggest that the proposed system is suitable for an objective assessment of posture and lower limb UPDRS tasks, also in a domestic environment, and then it could be the basis for the development of neuromonitoring and neurorehabilitation applications in a telemedicine framework.

**Author Contributions:** C.F. and R.N. designed and developed the system, analyzed the PD data and wrote the paper; A.C. and G.P. gave technical support on the development and contributed to review the paper; N.C. and V.C. provided the optoelectronic facilities and data; C.A., L.P. and A.M. designed and supervised the clinical experiment on PD subjects and assessed the patient performance.

**Funding:** This work was partially supported by VREHAB project, funded by the Italian Ministry of Health (RF-2009-1472190).

**Conflicts of Interest:** The authors declare no conflict of interest.
