This paper presents a novel method for feature extraction from electroencephalogram (EEG) signals using empirical mode decomposition (EMD). Its use is motivated by the fact that the EMD gives an effective time-frequency analysis of nonstationary signals. The intrinsic mode functions (IMF) obtained as a result of EMD give the decomposition of a signal according to its frequency components. We present the usage of upto third order temporal moments, and spectral features including spectral centroid, coefficient of variation and the spectral skew of the IMFs for feature extraction from EEG signals. These features are physiologically relevant given that the normal EEG signals have different temporal and spectral centroids, dispersions and symmetries when compared with the pathological EEG signals. The calculated features are fed into the standard support vector machine (SVM) for classification purposes. The performance of the proposed method is studied on a publicly available dataset which is designed to handle various classification problems including the identification of epilepsy patients and detection of seizures. Experiments show that good classification results are obtained using the proposed methodology for the classification of EEG signals. Our proposed method also compares favorably to other state-of-the-art feature extraction methods.

No abstract

BackgroundIt is estimated that 287,000 women worldwide die annually from pregnancy and childbirth-related conditions, and 6.9 million under-five children die each year, of which about 3 million are newborns. Most of these deaths occur in sub-Saharan Africa. The maternal health situation in Tanzania mainland and Zanzibar is similar to other sub-Saharan countries. This study assessed the availability, accessibility and quality of emergency obstetric care services and essential resources available for maternal and child health services in Zanzibar.MethodsFrom October and November 2012, a cross-sectional health facility survey was conducted in 79 health facilities in Zanzibar. The health facility tools developed by the Averting Maternal Death and Disability program were adapted for local use.ResultsOnly 7.6 % of the health facilities qualified as functioning basic EmONC (Emergency Obstetric and Neonatal Care) facilities and 9 % were comprehensive EmONC facilities. Twenty-eight percent were partially performing basic EmONC and the remaining 55.7 % were not providing EmONC. Neonatal resuscitation was performed in 80 % of the hospitals and only 17.4 % of the other health facilities that were surveyed. Based on World Health Organisation (WHO) criteria, the study revealed a gap of 20 % for minimum provision of EmONC facilities per 500,000 population. The met need at national level (proportion of women with major direct obstetric complications treated in a health facility providing EmONC) was only 33.1 % in the 12 months preceding the survey. The study found that there was limited availability of human resources in all visited health facilities, particularly for the higher cadres, as per Zanzibar minimum staff requirements.ConclusionThere is a need to strengthen human resource capacity at primary health facilities through training of health care providers to improve EmONC services, as well as provision of necessary equipment and supplies to reduce workload at the higher referral health facilities and increase geographic access.

No abstract

Objectives: As for stroke rehabilitation, brain-computer interfaces could potentially be used for inducing neural plasticity in patients with cerebral palsy by pairing movement intentions with relevant somatosensory feedback. Therefore, the aim of this study was to investigate if movement intentions from children with cerebral palsy can be detected from single-trial EEG. Moreover, different feature types and electrode setups were evaluated. Approach: Eight adolescents with cerebral palsy performed self-paced dorsiflexions of the ankle while nine channels of EEG were recorded. The EEG was divided into movement intention epochs and idle epochs. The data were pre-processed and temporal, spectral and template matching features were extracted and classified using a random forest classifier. The classification accuracy of the 2-class problem was used as an estimation of the detection performance. This analysis was repeated using a single EEG channel, a Large Laplacian filtered channel and nine channels. Results: A classification accuracy of ~70% was obtained using only a single channel. This increased to ~80% for the Laplacian filtered data, while ~75% of the data were correctly classified when using nine channels. In general, the highest accuracies were obtained using temporal features or using all of them combined. Significance: The results indicate that it is possible to detect movement intentions in patients with cerebral palsy; this may be used in the development of a brain-computer interface for motor rehabilitation of patients with cerebral palsy.

No abstract

Preventive healthcare requires continuous monitoring of the blood pressure (BP) of patients, which is not feasible using conventional methods. Photoplethysmogram (PPG) signals can be effectively used for this purpose as there is a physiological relation between the pulse width and BP and can be easily acquired using a wearable PPG sensor. However, developing real-time algorithms for wearable technology is a significant challenge due to various conflicting requirements such as high accuracy, computationally constrained devices, and limited power supply. In this paper, we propose a novel feature set for continuous, real-time identification of abnormal BP. This feature set is obtained by identifying the peaks and valleys in a PPG signal (using a peak detection algorithm), followed by the calculation of rising time, falling time and peak-to-peak distance. The histograms of these times are calculated to form a feature set that can be used for classification of PPG signals into one of the two classes: normal or abnormal BP. No public dataset is available for such study and therefore a prototype is developed to collect PPG signals alongside BP measurements. The proposed feature set shows very good performance with an overall accuracy of approximately 95%. Although the proposed feature set is effective, the significance of individual features varies greatly (validated using significance testing) which led us to perform weighted voting of features for classification by performing autoregressive modeling. Our experiments show that the simplest linear classifiers produce very good results indicating the strength of the proposed feature set. The weighted voting improves the results significantly, producing an overall accuracy of about 98%. Conclusively, the PPG signals can be effectively used to identify BP, and the proposed feature set is efficient and computationally feasible for implementation on standalone devices.

No abstract