Stijn Luca

Accelerometry-Based Home Monitoring for Detection of Nocturnal Hypermotor Seizures Based on Novelty Detection

Nocturnal home monitoring of epileptic children is often not feasible due to the cumbersome manner of seizure monitoring with the standard method of video/EEG-monitoring. We propose a method for hypermotor seizure detection based on accelerometers attached to the extremities. From the acceleration signals, multiple temporal, frequency, and wavelet-based features are extracted. After determining the features with the highest discriminative power, we classify movement events in epileptic and nonepileptic movements. This classification is only based on a nonparametric estimate of the probability density function of normal movements. Such approach allows us to build patient-specific models to classify movement data without the need for seizure data that are rarely available. If, in the test phase, the probability of a data point (event) is lower than a threshold, this event is considered to be an epileptic seizure; otherwise, it is considered as a normal nocturnal movement event. The mean performance over seven patients gives a sensitivity of 95.24% and a positive predictive value of 60.04%. However, there is a noticeable interpatient difference.

Detecting rare events using extreme value statistics applied to epileptic convulsions in children

OBJECTIVE Nocturnal home monitoring of epileptic children is often not feasible due to the cumbersome manner of seizure detection with the standard method of video electroencephalography monitoring. The goal of this paper is to propose a method for hypermotor seizure detection based on accelerometers that are attached to the extremities. METHODS Supervised methods that are commonly used in literature need annotation of data and hence require expert (neurologist) interaction resulting in a substantial cost. In this paper an unsupervised method is proposed that uses extreme value statistics and seizure detection based on a model of normal behavior that is estimated using all recorded and unlabeled data. In this way the expensive interaction can be avoided. RESULTS When applying this method to a labeled dataset, acquired from 7 patients, all hypermotor seizures are detected in 5 of the 7 patients with an average positive predictive value (PPV) of 53%. For evaluating the performance on an unlabeled dataset, seizure events are presented to the system as normal movement events. Since hypermotor seizures are rare compared to normal movements, the very few abnormal events have a negligible effect on the quality of the model. In this way, it was possible to evaluate the system for 3 of the 7 patients when 3% of the training set was composed of seizure events. This resulted in sensitivity scores of 80%, 22% and 90% and a PPV of 89%, 21% and 44% respectively. These scores are comparable with a state-of-the-art supervised machine learning based approach which requires a labeled dataset. CONCLUSIONS A person-dependent epileptic seizure detection method has been designed that requires little human interaction. In contrast to traditional machine learning approaches, the imbalance of the dataset does not cause substantial difficulties.

Developing a system that can automatically detect health changes using transfer times of older adults

BackgroundAs gait speed and transfer times are considered to be an important measure of functional ability in older adults, several systems are currently being researched to measure this parameter in the home environment of older adults. The data resulting from these systems, however, still needs to be reviewed by healthcare workers which is a time-consuming process.MethodsThis paper presents a system that employs statistical process control techniques (SPC) to automatically detect both positive and negative trends in transfer times. Several SPC techniques, Tabular cumulative sum (CUSUM) chart, Standardized CUSUM and Exponentially Weighted Moving Average (EWMA) chart were evaluated. The best performing method was further optimized for the desired application. After this, it was validated on both simulated data and real-life data.ResultsThe best performing method was the Exponentially Weighted Moving Average control chart with the use of rational subgroups and a reinitialization after three alarm days. The results from the simulated data showed that positive and negative trends are detected within 14 days after the start of the trend when a trend is 28 days long. When the transition period is shorter, the number of days before an alert is triggered also diminishes. If for instance an abrupt change is present in the transfer time an alert is triggered within two days after this change. On average, only one false alarm is triggered every five weeks. The results from the real-life dataset confirm those of the simulated dataset.ConclusionsThe system presented in this paper is able to detect both positive and negative trends in the transfer times of older adults, therefore automatically triggering an alarm when changes in transfer times occur. These changes can be gradual as well as abrupt.

Automatic detection of health changes using statistical process control techniques on measured transfer times of elderly.

It has been shown that gait speed and transfer times are good measures of functional ability in elderly. However, data currently acquired by systems that measure either gait speed or transfer times in the homes of elderly people require manual reviewing by healthcare workers. This reviewing process is time-consuming. To alleviate this burden, this paper proposes the use of statistical process control methods to automatically detect both positive and negative changes in transfer times. Three SPC techniques: tabular CUSUM, standardized CUSUM and EWMA, known for their ability to detect small shifts in the data, are evaluated on simulated transfer times. This analysis shows that EWMA is the best-suited method with a detection accuracy of 82% and an average detection time of 9.64 days.

Modified chain sampling plans for lot inspection by variables and attributes

ABSTRACT The purpose of acceptance sampling is to develop decision rules to accept or reject production lots based on sample data. When testing is destructive or expensive, dependent sampling procedures cumulate results from several preceding lots. This chaining of past lot results reduces the required size of the samples. A large part of these procedures only chain past lot results when defects are found in the current sample. However, such selective use of past lot results only achieves a limited reduction of sample sizes. In this article, a modified approach for chaining past lot results is proposed that is less selective in its use of quality history and, as a result, requires a smaller sample size than the one required for commonly used dependent sampling procedures, such as multiple dependent sampling plans and chain sampling plans of Dodge. The proposed plans are applicable for inspection by attributes and inspection by variables. Several properties of their operating characteristic-curves are derived, and search procedures are given to select such modified chain sampling plans by using the two-point method.

Anomaly Detection Using the Poisson Process Limit for Extremes

Anomaly detection starts from a model of normal behavior and classifies departures from this model as anomalies. This paper introduces a statistical non-parametric approach for anomaly detection that is based on a multivariate extension of the Poisson point process model for univariate extremes. The method is demonstrated on both a synthetic and a real-world data set, the latter being an unbalanced data set of acceleration data collected from movements of 7 pediatric patients suffering from epilepsy that is previously studied in [1]. The positive predictive values could be improved with an increase up to 12.9% (and a mean of 7%) while the sensitivity scores stayed unaltered. The proposed method was also shown to outperform an one-class SVM classifier. Because the Poisson point process model of extremes is able to combine information on the number of excesses over a fixed threshold with that on the excess values, a powerful model to detect anomalies is obtained that can be of high value in many applications.

Point process models for novelty detection on spatial point patterns and their extremes

Abstract Novelty detection is a particular example of pattern recognition identifying patterns that departure from some model of “normal behaviour”. The classification of point patterns is considered that are defined as sets of N observations of a multivariate random variable X and where the value N follows a discrete stochastic distribution. The use of point process models is introduced that allow us to describe the length N as well as the geometrical configuration in data space of such patterns. It is shown that such infinite dimensional study can be translated into a one-dimensional study that is analytically tractable for a multivariate Gaussian distribution. Moreover, for other multivariate distributions, an analytic approximation is obtained, by the use of extreme value theory, to model point patterns that occur in low-density regions as defined by X . The proposed models are demonstrated on synthetic and real-world data sets.

Prediction of clinical foot characteristics using quantitative features from different measurement set-ups

We measured 77 healthy subjects without major foot deformities. They were all clinically assessed by 10 experts (orthopedic technologists, podiatrists, and one orthopedic surgeon), hence a total of 770 assessments. Furthermore, an anamnesis was conducted, and gait of all subjects was quantitatively measured using three-dimensional (3D) motion analysis (Codamotion), dynamic pressure plate (RSScan International), a dynamic 3D scanner (ViALUX), and a force plate (AMTI). To identify those clinical characteristics, which are robust over the different experts, we conducted a 2agreement weighted kappa analysis which is an extension of Cohen’s kappa for multiple raters (Warrens, 2012). Furthermore, we included both the popularity and the discriminative power of a characteristic (i.e. how many experts scored it and how diverse are the scores, respectively.). We included these last two elements because if either popularity or discriminative power are low, we cannot say much about a certain feature, e.g. if it is evaluated by only one or two experts, or if all subjects get the same score. In a second part, we used the quantitatively extracted features (from the pressure plate, 3D motion analysis, dynamic 3D scanner, and force plate) to predict the average expert scores, for each clinical characteristic individually. To determine the best feature subset, we carried out a feature selection using the Lasso technique in a 10-fold cross validation. The feature subset was then fed to a support vector machine (SVM) classifier which trained a prediction model using a leave-one-out crossvalidation. Finally, from these data we can give an indication which hardware is best to predict foot characteristics. To this end, we built the SVM model only including features from one or a limited set of measurement equipment. In this abstract, we highlight three cases: prediction of the resting calcaneal stance position (RCSP), pressure of the midfoot during stance, and the ratio of the forefoot/heel width.