Narrow Results

The development of efficient stroke-detection methods is of significant importance in today's society due to the effects and impact of stroke on health and economy worldwide. This study focuses on Human Activity Recognition (HAR), which is a key component in developing an early stroke-diagnosis tool. An overview of the proposed global approach able to discriminate normal resting from stroke-related paralysis is detailed. The main contributions include an extension of the Genetic Fuzzy Finite State Machine (GFFSM) method and a new hybrid feature selection (FS) algorithm involving Principal Component Analysis (PCA) and a voting scheme putting the cross-validation results together. Experimental results show that the proposed approach is a well-performing HAR tool that can be successfully embedded in devices.

The identification and the modeling of epilepsy convulsions during everyday life using wearable devices would enhance patient anamnesis and monitoring. The psychology of the epilepsy patient penalizes the use of user-driven modeling, which means that the probability of identifying convulsions is driven through generalized models. Focusing on clonic convulsions, this pre-clinical study proposes a method for generating a type of model that can evaluate the generalization capabilities. A realistic experimentation with healthy participants is performed, each with a single 3D accelerometer placed on the most affected wrist. Unlike similar studies reported in the literature, this proposal makes use of $5\times 2$ cross-validation scheme, in order to evaluate the generalization capabilities of the models. Event-based error measurements are proposed instead of classification-error measurements, to evaluate the generalization capabilities of the model, and Fuzzy Systems are proposed as the generalization modeling technique. Using this method, the experimentation compares the most common solutions in the literature, such as Support Vector Machines, $k$-Nearest Neighbors, Decision Trees and Fuzzy Systems. The event-based error measurement system records the results, penalizing those models that raise false alarms. The results showed the good generalization capabilities of Fuzzy Systems.

Wearable devices have promoted the application of Human Activity Recognition to the development of techniques for the assessment or diagnosing of illnesses and seizures, among other applications. For instance, the use of tri-axial accelerometry (3DACM) to detect abnormal and sudden movements has been introduced in the epileptic seizure recognition. In a previous research, Fuzzy Rule Based Classifiers (FRBC) have been found valid for the detection of epileptic convulsions; however, Ant Colony Systems learned FRBC performed with a high variability depending on the training data. In this study, we cope with this problem by the selection of a suitable partitioning method that has been extended to generate Fuzzy partitions. The comparison with the previous obtained results shows the fuzzy partitioning does not improve the overall performance in terms of error but highly reduces the variability in the performance of the obtained models, which allows us to obtain general models.