Various methods to determine the onset of the electromyographic activity which occurs in response to a stimulus have been discussed in the literature over the last decade. Due to the stochastic characteristic of the surface electromyogram (SEMG), onset detection is a challenging task, especially in weak SEMG responses. The performance of the onset detection methods were tested, mostly by comparing their automated onset estimations to the manually determined onsets found by well-trained SEMG examiners. But a systematic comparison between methods, which reveals the benefits and the drawbacks of each method compared to the other ones and shows the specific dependence of the detection accuracy on signal parameters, is still lacking. In this paper, several classical threshold-based approaches as well as some statistically optimized algorithms were tested on large samples of simulated SEMG data with well-known signal parameters. Rating between methods is performed by comparing their performance to that of a statistically optimal maximum likelihood estimator which serves as reference method. In addition, performance was evaluated on real SEMG data obtained in a reaction time experiment. Results indicate that detection behavior strongly depends on SEMG parameters, such as onset rise time, signal-to-noise ratio or background activity level. It is shown that some of the threshold-based signal-power-estimation procedures are very sensitive to signal parameters, whereas statistically optimized algorithms are generally more robust

No abstract

Investigation of the human motor system frequently requires precise determination of the motor response onset indicating the time of movement initiation (e.g., in reaction time experiments). This paper presents a new model-based algorithm for computerized response onset detection in kinematic signals (e.g., joint angle). The response onset is identified as an abrupt change in the (time-varying) parameters of a statistical process model adapted to the measured signal. The accuracy of the algorithm is assessed by statistical simulations, and the performance of the method is compared to the performance of conventional onset detection methods using simulated as well as real kinematic signals. Results show that onset detection can substantially be improved by including a priori knowledge on the physiological background of the measured signals to the decision process.

We describe the first steps in the development of a wearable measurement device for measuring a subject's three-dimensional acceleration. The ultimate aim is a standard measurement instrument integrated in a belt buckle that allows objective evaluation of treatment and rehabilitation measures in patients, in particular for disabling chronic diseases such as multiple sclerosis. In a first step we combined standard hardware elements to record test data from healthy volunteers. We then developed algorithms to automatically distinguish between different stages of activity, such as jogging, walking, lying, standing and sitting, and to detect and count steps. Distinction between standing and sitting is the most difficult to accomplish. As a first validation, we calculated the distance traveled from data of 17 experiments and a total of 4.5 h, for which one proband was walking and running for a known distance, and compared the results with two commercially available pedometers. We could show that the relative error for the ActiBelt is only half of that for the two pedometers. Apart from developing much smaller, robust and integrated hardware, we describe ideas on how to develop algorithms that allow extraction of a "baseline step pattern" in analogy to baseline ECG to define and detect clinically relevant deviations.

No abstract