Publications

In the context of the occidental population aging, new preventive approaches must be developed to reduce the disability related to brain stroke in the elderly. The stroke susceptibility assessment based on the analysis of human movements is one of the potential avenues needing investigation. As a first step in this direction, this paper reports results on the relationship linking the most important stroke risk factors to some characteristics of human movement. Various features were extracted using the Sigma-Lognormal model on 1440 stereotypical triangular movements performed by 120 subjects having different health conditions. These features were combined through a linear modeling to maximize the predictability of presence of stroke risk factors in the studied cohort. The receiver operating characteristic (ROC) curve and the area under this curve (AUC) were used to evaluate the clinical significance of this relationship. Using only the information derived from the movements, the six tested risk factors (cardiac problems, diabetes mellitus, hypercholesterolemia, hypertension, obesity, and cigarette smoking) can be predicted with an AUC ranging from .68 to .82.

The main goal of this work is to determine whether a computer mouse can be used as a low-cost device for the acquisition of two-dimensional human movement velocity signals in the context of psychophysical studies and biomedical applications. A comprehensive overview of the related literature is presented, and the problem of characterizing mouse movement acquisition is analyzed and discussed. Then, the quality of velocity signals acquired with this kind of device is measured on horizontal oscillatory movements by comparing the mouse data to the signals acquired simultaneously by a video motion tracking system and a digitizing tablet. A synthesis of the information gathered in this work indicates that the computer mouse can be used for the reliable acquisition of biosignals in the context of human movement studies, particularly for many applications dealing with the velocity of the end effector of the upper limb. This paper concludes by discussing the possibilities and limitations of such use.

In this paper, 14 healthy subjects in two age groups have produced rapid handwriting strokes with a direction reversal. The delta-lognormal model was used to obtain a detailed description of the velocity of these movements and of the neuromuscular synergy that produces them. This modeling also allowed the derivation of new hypothesis on the nature of the slowing effect due to aging (i.e., a direct effect or a coping strategy) and on its repartition on the different steps of the movement production (i.e., its preparation versus its execution). Our analysis revealed a substantial increase of neuromuscular response delays and a decrease of the command amplitudes with age. For the parameters that show a significant decrease in performance, the agonist and antagonist systems were affected similarly. In addition, we observed that the age has a proportional effect on the various time characteristics of the movements and that even in the case of a significant slowing down of the neuromuscular systems, the elderly can still achieve optimal movement responses, characterized by the use of a single delta-lognormal primitive. This performance might be related to the preservation of some movement timing properties and relationships between the agonist and the antagonist neuromuscular systems.

In this paper, the existence of oscillations for a class of recurrent neural networks with time delays between neural interconnections is investigated. By using the fixed point theory and Liapunov functional, we prove that a recurrent neural network might have a unique equilibrium point which is unstable. This particular type of instability, combined with the boundedness of the solutions of the system, will force the network to generate a permanent oscillation. Some necessary and sufficient conditions for these oscillations are obtained. Simple and practical criteria for fixing the range of parameters in this network are also derived. Typical simulation examples are presented.