Investigating motor imagery tasks by their neural effects - A case study.

Motor imagery, one of the first investigated neural process for Brain-Computer Interfaces (BCIs) still provides a great challenge nowadays. Aiming a better and more accurate control, multiple researches have been conducted by the scientific community. Nevertheless, there is still no robust and confident application developed. In order to augment the potential referring to motor imagery, and to attract user's interest, we propose multiple motor imagery tasks in combination with different visual or auditory stimuli. We use multi-class classification for discrimination and we observe confident classification performance for the task related to user's background.

Abdominal signals: denoising by application of the event synchronuous canceller.

The electrical activity recorded on the abdomen during pregnancy and labor (abdominal signals, ADS) contains vital information about the health state of both mother and fetus. The most important signal related with the health of fetus, extensively studied by now, is the fetal ECG (fECG) which allows physicians to examine the evolution of the fetus and to identify possible heart diseases of fetus. The movement of the fetus which shows up in the electrohysterogram (EHG), extracted from ADS, indicates a normal pregnancy. This paper presents a method of removing the maternal ECG (mECG) from ADS in order extract the EHG and to obtain information about the uterine contractions and fetal movements. After removing the mECG and the fECG, the ADS, filtered in the frequency range of the uterine activity, is further used to predict labor or preterm labor. The ADS, cleaned from the mECG allows also the analysis of the fECG for fetal monitoring.