This group is trying to enable wireless communication between several types of a miniaturized bio-sensors and bio-central devices worn at the human body. We aim to develop a system integration technology for reliable bio-signal/bio-information processing and establish wireless networks with bionic application devices in healthcare domain.
A brain-computer interface (BCI) is a system that uses brain signals to provide a non-muscular communication channel, particularly for individuals with severe neuromuscular disabilities. The electroencephalogram (EEG) signals for the BCI system suffer from high noise levels due to low conductivity of the human skull. Regarding the property of EEG signals our objective is to develop various feature extractions and classifications for the BCI system.
Electro-encephalogram (EEG) refers to the measurement of electrical signals from the scalp that result from the neuronal activity of the brain. These measurements are usually done through the standard wet Ag/AgCl electrodes. This approach provides high quality signals but it can cause some discomfort as it requires the use of conductive gels as well as the scalp preparation Also, standard "dry" electrodes have high skin-electrode impedance that degrades the already low signal level of the EEG signals.
We aim to develop the dry EEG electrode that has high sensitivity and mechanical sturdiness. Especially, we focus on the approach minimizes the skin-electrode air gap based on the curvature of the scalp, provides significantly lower impedance.
Fall detection systems are meant to send an alert message for help when a fall event occurs which helps to avoid long-lie situations after the fall allowing timely interventions, reduce the fear of falling and ultimately improves the quality of life. Fall risk prediction tools try to identify individuals who are more likely to fall in near future, so that fall prevention intervention can be provided prior to actual falls.
Falls often result from interacting factors which will be different for each patient. The existing clinical fall-risk assessment approaches require expert supervision, have lack of objectiveness, offers high healthcare costs, ineffective use of time and resources and not suitable for long term monitoring of subject. Wireless wearable sensors offer a promising approach for real-time and continuous monitoring of falls and risk of falling during daily activities. We are focusing upon efficient fall-risk assessment via monitoring and analyzing of daily activities where the major challenge lies in finding optimal Sensor placement sites with useful features extraction and to match predictive variables/features with specific risk factors. We are also involved in activity recognition and falls detection where the key challenge is to classify between everyday activities and falls to reduce false alarms. Along with these, energy consumption, ease of use and convenience of older adults is also taken into consideration.