The first level of data aggregation is the patient’s equipment data acquisition sub-layer, which in close collaboration with the ARMOR Middleware, provides access towards the upper modules and vice versa. The role of these sub-layers is crucial since they comprise the gateway point between sensor hardware equipment and software application functional modules. On one hand support is provided for all types of data and sensors that are utilized in the context ARMOR and on the other hand a wide range of services is provided in order to support the depicted functional modules. Furthermore, they comprise the communication bus among the aforementioned modules thus emphasizing on the criticality of these sub layers.

ARMOR Information server is a key ingredient of the ARMOR system. It hosts the models derived from extended research effort, which intend for the multi-parametric data analysis. As indicated in Fig. 1.1, the development of new methods and tools regarding online/offline analysis, data fusion, real time processing, data profiling, cleaning and transformation that enable reliable, accurate and economical diagnosis through models are integrated in this module. Additionally, hardware-based acceleration modules have been incorporated in order to facilitate the demanding task of data analysis. However, such functionality requires close collaboration with patient’s data stored in Electronic Health Record (EHR) database (which is detailed in Chap. 11), a secure database which comprises the second functional module of the ARMOR system where all patient data are stored in an anonymized, yet efficient to process way.

The third functional module of ARMOR platform is ARMOR Application Server which essentially provides access to the ARMOR system by the specific personnel (i.e. patient, medical stuff, caregivers) through a wide range of user interfaces. Such interfaces include multi-parametric data processing results visualization, EHR access and personal tracking and vital signs information.

The ARMOR system provides also novel functionalities for managing and analyzing both new and already acquired multimodal data of epileptic patients (e.g. EEG, EMG, ECG, PET, sensor data, genetic data). As it will be presented in Chap. 14, as a first step some of the existing well-established techniques for data cleaning, integration and transformation were reused and enhanced for the pre-processing of multi-parametric data. In order to deal with the large amount of ARMOR’s data, new contributions have been made in data and dimensionality reduction, involving feature selection and extraction methodologies together with new representation techniques for time series data. ARMOR’s data, such as logs of all events, recorded values from sensors and other metrics that are monitored, together with personalized patient health profiles, medical information for diseases, medication, etc., have been organized in a data management system.

In order to achieve more efficient and accurate inferences, data fusion techniques have also been explored. Existing fusion techniques have been extended and new ones have been developed for the integration and fusion of information from sensors, already stored data about the patients that the clinical collaborators of ARMOR project collected or had already available, along with the personalized patient health profiles. Novel real time (online) analysis methods for multi-parametric streaming data that have also been developed, which aim at detecting signals beyond the limits, identify seizure premonitory signs, and discover typical patterns of activity followed by seizures and atypical patterns of activity, based on the models created. All analysis and emergency alert mechanisms offered by ARMOR platform are based on a personalized model according to the patient’s health profile. Moreover, new decision support tools for advising the medical professional, and the patient, triggering an alarm, and detecting emergency situations have been developed. Chapters 10 and 11 provide an in depth analysis of the corresponding aspects.