Data Mining - Concepts and Techniques.

Falls are a serious public health problem and possibly life threatening for people in fall risk groups. We develop an automated fall detection system with wearable motion sensor units fitted to the subjects' body at six different positions. Each unit comprises three tri-axial devices (accelerometer, gyroscope, and magnetometer/compass). Fourteen volunteers perform a standardized set of movements including 20 voluntary falls and 16 activities of daily living (ADLs), resulting in a large dataset with 2520 trials. To reduce the computational complexity of training and testing the classifiers, we focus on the raw data for each sensor in a 4 s time window around the point of peak total acceleration of the waist sensor, and then perform feature extraction and reduction. Most earlier studies on fall detection employ rule-based approaches that rely on simple thresholding of the sensor outputs. We successfully distinguish falls from ADLs using six machine learning techniques (classifiers): the k-nearest neighbor (k-NN) classifier, least squares method (LSM), support vector machines (SVM), Bayesian decision making (BDM), dynamic time warping (DTW), and artificial neural networks (ANNs). We compare the performance and the computational complexity of the classifiers and achieve the best results with the k-NN classifier and LSM, with sensitivity, specificity, and accuracy all above 99%. These classifiers also have acceptable computational requirements for training and testing. Our approach would be applicable in real-world scenarios where data records of indeterminate length, containing multiple activities in sequence, are recorded.

/pdf/detecting-falls-with-wearable-sensors-using-machine-learning-3hqp504tlw.pdf

Detecting falls with wearable sensors using machine learning techniques.

Why peer discussion improves student performance on in-class concept questions

Abstract. This paper introduces a new frequency-domain approach to describe the relationships (direction of information flow) between multivariate time series based on the decomposition of multivariate partial coherences computed from multivariate autoregressive models. We discuss its application and compare its performance to other approaches to the problem of determining neural structure relations from the simultaneous measurement of neural electrophysiological signals. The new concept is shown to reflect a frequency-domain representation of the concept of Granger causality.

Partial directed coherence: a new concept in neural structure determination

This perspective article will discuss the potential role of body-worn movement monitors for balance and gait assessment and treatment in rehabilitation. Recent advances in inexpensive, wireless sensor technology and smart devices are resulting in an explosion of miniature, portable sensors that can quickly and accurately quantify body motion. Practical and useful movement monitoring systems are now becoming available. It is critical that therapists understand the potential advantages and limitations of such emerging technology. One important advantage of obtaining objective measures of balance and gait from body-worn sensors is impairment-level metrics characterizing how and why functional performance of balance and gait activities are impaired. Therapy can then be focused on the specific physiological reasons for difficulty in walking or balancing during specific tasks. A second advantage of using technology to measure balance and gait behavior is the increased sensitivity of the balance and gait measures to document mild disability and change with rehabilitation. A third advantage of measuring movement, such as postural sway and gait characteristics, with body-worn sensors is the opportunity for immediate biofeedback provided to patients that can focus attention and enhance performance. In the future, body-worn sensors may allow therapists to perform telerehabilitation to monitor compliance with home exercise programs and the quality of their natural mobility in the community. Therapists need technological systems that are quick to use and provide actionable information and useful reports for their patients and referring physicians. Therapists should look for systems that provide measures that have been validated with respect to gold standard accuracy and to clinically relevant outcomes such as fall risk and severity of disability.

Role of body-worn movement monitor technology for balance and gait rehabilitation.

We examined falling risk among elderly using a wearable inertial sensor, which combines accelerometer and gyrosensors devices, applied during the Timed Up and Go (TUG) test. Subjects were categorised into two groups as low fall risk and high fall risk with 13.5 s duration taken to complete the TUG test as the threshold between them. One sensor was attached at the subject's waist dorsally, while acceleration and gyrosensor signals in three directions were extracted during the test. The analysis was carried out in phases: sit-bend, bend-stand, walking, turning, stand-bend and bend-sit. Comparisons between the two groups showed that time parameters along with root mean square (RMS) value, amplitude and other parameters could reveal the activities in each phase. Classification using RMS value of angular velocity parameters for sit-stand phase, RMS value of acceleration for walking phase and amplitude of angular velocity signal for turning phase along with time parameters suggests that this is an improved method in evaluating fall risk, which promises benefits in terms of improvement of elderly quality of life.

Quantitative analysis of fall risk using TUG test

Body temperature of the baby is an important element that will inform the actual health condition of a baby. Parents and caregiver occasionally do not aware of drastic increase in body temperature in a short period of time and febrile seizure may happen that could leads to epilepsy. In Malaysia, device that could monitor the infant body's temperature is already available. However, the device could not continuously be used for a long hours and create discomfort to the babies due to its size of devices. Therefore, a small, lightweight device that continuously monitors the body temperature and comfortably used by baby is developed. It directly helps parents by alerting them whenever the baby's body temperature increased higher than normal a degree. This system monitors the vital parameter which is the body temperature by using a wearable sensor. The information then transferred to their parents through a wireless network. The system is extended for interfacing with the mobile phones to enable remote monitoring. Architecture of the system consists of a wearable sensor for monitoring the vital parameter and a sound buzzer where all of the component be controlled by a single microcontroller, the ESPresso Lite V2.0 based on ESP8266 and supplied by the lithium ion polymer battery. Even though the system is more focusing on temperature monitoring only, it can be further expanded or upgrade to monitor other vital parameter such as heart rate, oxygen saturation, respiration rate or any other parameter.

IoT (Internet of Things) Based Infant Body Temperature Monitoring

Modeling Brain Activities during Reading Working Memory Task: Comparison between Reciting Quran and Reading Book☆

We performed a quantitative analysis of the fall-risk assessment test using a wearable inertia sensor focusing on two tests: the time up and go (TUG) test and the four square step test (FSST). These tests consist of various daily activities, such as sitting, standing, walking, stepping, and turning. The TUG test was performed by subjects at low and high fall risk, while FSST was performed by healthy elderly and hemiplegic patients with high fall risk. In general, the total performance time of activities was evaluated. Clinically, it is important to evaluate each activity for further training and management. The wearable sensor consisted of an accelerometer and angular velocity sensor. The angular velocity and angle of pitch direction were used for TUG evaluation, and those in the pitch and yaw directions at the thigh were used for FSST. Using the threshold of the angular velocity signal, we classified the phase corresponding to each activity. We then observed the characteristics of each activity and recommended suitable training and management. The wearable sensor can be used for more detailed evaluation in fall risk management. The wearable sensor can be used more detailed evaluation for fall-risk management test.

Quantitative analysis of the fall-risk assessment test with wearable inertia sensors

Solat adalah salah satu cara umat Islam untuk mengabdikan diri kepada Allah. Kajian ini bertujuan untuk mengkaji kesan selepas solat terhadap otak dengan menggunakan Electroencephalograph(EEG) dan kegiatan saraf autonomik. Sebanyak 20 sampel isyarat EEG daripada 5 responden (lelaki berumur di antara 22 hingga 23 tahun) dipilih, dan mereka bebas daripada masalah jantung, paru–paru, metabolik dan penyakit lain yang boleh menggangu kegiatan saraf autonomik. Analisa dilakukan dengan membuat perbandingan di antara sampel EEG bagi dua aktiviti berbeza iaitu aktiviti selepas solat dan aktiviti selepas mendengar muzik menggunakan kaedah analisa prinsip komponen (PCA) dan transformasi fourier (FFT). Hasil analisa ke atas isyarat mendapati amplitud tertinggi pada agihan spektrum kuasa adalah pada jalur gamma bagi EEG berbanding jalur–jalur lain (delta, theta, alpha dan beta) bagi kedua–dua aktiviti. Kajian juga menunjukkan kelima–lima subjek mempunyai jalur gamma yang tinggi selepas solat berbanding selepas mendengar muzik. Keputusan ini mencadangkan, apabila seseorang muslim melakukan solat, mereka akan melupakan seketika masalah mereka dan khusyuk mengadap Allah. Ini dapat membantu memulihkan tekanan yang dialami mereka untuk beberapa ketika, seterusnya lebih bersedia untuk menghadapi segala permasalahan dunia selepas solat. 
				
				Kata kunci: Solat; electroencephalograph (EEG); isyarat gelombang otak

				Salat is one of the ways for the Muslims to worship ‘Allah’ (God). This study investigates the brainwave signal after salat using Electroencephalograph (EEG) signal analysis and autonomic nervous activity. Twenty samples of EEG signal from five respondents (males with age range from 22 to 23 years old) were selected, and they were all free from cardiac, pulmonary, metabolic and other disease that would cause autonomic nervous system dysfunction. The analysis of the acquired EEG signal is done by comparing the samples of EEG signal between the activities after salat and after listening to the music using Principal Component Analysis (PCA) and Fast Fourier Transform (FFT) method. The highest amplitude of the power spectrum distribution was observed in the gamma band on EEG compared to other bands (delta, theta, alpha and beta) for both activities. It shows that these five subjects have a high gamma band after salat compared to after listening to music. It is suggested that when Muslim perform salat, they tend to forget all the worldly matters and thinks they are standing in front of the ‘Allah’. This eventually help them to break their stress for a few times, hence reducing its intensity to facing the worlds’ matter after salat. 
				
				Key words: Solat; electroencephalograph (EEG); brainwave signal

N. A. Zakaria

Papers

Quantitative analysis of fall risk using TUG test

IoT (Internet of Things) Based Infant Body Temperature Monitoring

Modeling Brain Activities during Reading Working Memory Task: Comparison between Reciting Quran and Reading Book☆

Quantitative analysis of the fall-risk assessment test with wearable inertia sensors

Salat and brainwave signal analysis