Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2007
Additive and multiplicative noise reduction by back propagation neural network.
A novel filter is proposed by applying back propagation neural network (BPNN) ensemble where the noisy signal and the reference one are the same. The neural network(NN) ensemble filter not only well reduces additive and multiplicative white noise inside signals, but also preserves signals' characteristics. ⋯ Furthermore, it is presented of the relationship between noise reduction and bandwidth of noises. The performance of the NN ensemble filter is demonstrated in computer simulations and actual electroencephalogram (EEG) signals processing.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2007
Development of real-time motion artifact reduction algorithm for a wearable photoplethysmography.
This paper presents a motion artifact reduction algorithm for a real-time, wireless and wearable photoplethysmography (PPG) device for measuring heart beats. A wearable finger band PPG device consists of a 3-axis accelerometer, infrared LED, photo diode, a microprocessor and wireless module. Sources of the motion artifacts were investigated from the hand motions, through computing the correlations between the three directional finger motions and distorted PPG signals. ⋯ NLMS (Normalized Least Mean Square) adaptive filter (4th order) was employed in the algorithm. As a result, the signals' distortion rates were reduced from 52.34% to 3.53%, at frequencies between 1 and 2.5 Hz, which representing daily motions such walking and jogging. The wearable health monitoring device equipped with the motion artifact reduction algorithm can be integrated as a terminal in a so-called ubiquitous healthcare system, which provides a continuous health monitoring without interrupting a daily life.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2007
Modeling state entropy of the EEG and auditory evoked potentials: hypnotic and analgesic interactions.
Because of the complexity of raw electroencephalogram (EEG), for the anesthesiologist it is very difficult to evaluate the patient's hypnosis state. Because of this, several depth of anesthesia monitors have been developed, and are in current use at the operating room (OR). These monitors convert the information supplied by the EEG or derived signals into a simple, easy to understand index. ⋯ Hypnotic and analgesic drugs interact in different ways throughout the anaesthesia stages. The models obtained captured the different dynamic interaction of drugs, during the induction and maintenance phases, demonstrating that the model must have incorporated all this information in order to perform satisfactorily. Other information like haemodynamic variables might be included in the search for the optimum model.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2007
Real-time development of patient-specific alarm algorithms for critical care.
The state-of-the-art monitoring systems for critical care measure vital signs and generate alerts based on the logic of general patient population models, but they lack the capabilities of accurately correlating physiological data with clinical events and of adapting to individual patient's characteristics that do not fit the population models. This research examines the feasibility of developing patient-specific alarm algorithms in real time at the bedside and evaluates the potential of these algorithms in helping improve patient monitoring. Modular components that facilitate real-time development of alarm algorithms were added to a system that simultaneously collects physiological data and clinical annotations at the bedside. ⋯ The performance of patient-specific alarm algorithms improved as training data increased. Neural networks with eight hours of training data on average achieved a sensitivity of 0.96, a specificity of 0.99, a positive predictive value of 0.79, and an accuracy of 0.99; these figures were 0.84, 0.98, 0.72, and 0.98 respectively for the classification trees. These results suggest that real-time development of patient-specific alarm algorithms is feasible using machine learning techniques.
-
Conf Proc IEEE Eng Med Biol Soc · Jan 2007
Intelligent monitoring of critical pathological events during anesthesia.
Expert algorithms in the field of intelligent patient monitoring have rapidly revolutionized patient care thereby improving patient safety. Patient monitoring during anesthesia requires cautious attention by anesthetists who are monitoring many modalities, diagnosing clinically critical events and performing patient management tasks simultaneously. The mishaps that occur during day-to-day anesthesia causing disastrous errors in anesthesia administration were classified and studied by Reason [1]. ⋯ When detecting absolute hypovolaemia (AHV), moderate level of agreement was observed between RT-SAAM and the human expert (anesthetist) during surgical procedures. RT-SAAM is a clinically useful diagnostic tool which can be easily modified for diagnosing additional critical pathological events like relative hypovolaemia, fall in cardiac output, sympathetic response and malignant hyperpyrexia during surgical procedures. RT-SAAM is currently being tested at the Auckland City Hospital with ethical approval from the local ethics committees.