Smart Health Systems

We aim at the design and development of smart tools/systems, which incorporate sensing, computing and communication technologies, software/hardware modeling and system architectures towards personalized, pervasive, participatory, predictive and preventive healthcare. Our research work in the field has resulted in several products.

Semantic Annotation of Medical Images

Design and development of complete online systems which allow medical experts to annotate medical images with formality and efficiently perform complex search queries. The ontologies gain access to the essential ontological framework (FMA, ICD-10) and provide additional information, extracted from DICOM files, about the image and the patient.

SMARTDIAB: A communication platform for tele-monitoring and management of patients with Type 1 Diabetes

The platform enables health care professionals to monitor the health evolution of their patients, and assess the available information regarding blood glucose levels, insulin intake, diet, physical activity etc. and makes recommendations about their treatments (insulin scheme and doses). The provided services are based on state-of-the-art technologies and infrastructures including Wireless Personal Area Network (WPAN) and mobile communications networks.

CAROTID: Diagnosis of Carotid Atherosclerosis from B-mode Ultrasound Images

A modular software system designed to assist the interpretation of B-Mode ultrasound images of carotid artery has been developed. The system allows image pre-processing, manual and automatic segmentation of regions of interest, estimation of motion from sequences of images, extraction of texture features, statistical analysis and classification of the extracted features.

ANALYSIS: Computer Aided Diagnosis of Liver Lesions

A computer system that aids the diagnosis of focal liver lesions based on CT imaging data, has been developed. The system provides tools for the segmentation and the 3D-visulaization of Regions of Interest. Texture features along with Artificial Neural Networks are used in order to classify the liver lesions. Moreover, a system able to classify liver lesions form Ultrasound Images has been developed based on the use of fractal dimension and fuzzy logic methods.

GALENOS: Virtual Simulation and Treatment Planning in Clinical Radio-oncology

The software integrates all the steps needed during the treatment planning at a workstation, where the patient is virtually available via his/her CT scans. It supports operations of virtual CT, virtual simulator, treatment planning as well as navigation between them. Advanced techniques for the optimization of 3D treatment planning are available and telematic cooperation between remote experts is enabled.

EPILDA: Efficient Management and Analysis of Epileptic Patients' Data

The EPILDA (EPILepsy DAtabase) system is a wide-scale web-based data management and analysis tool aiming at assisting the efficient management and analysis of epidemiological, clinical, paraclinical, pre-surgical evaluation and therapeutic data concerning epilepsy. The EPILDA system is designed in a patient-oriented manner, enabling the user to easily navigate through the data management options and facilitating handling of tools for data analysis. The design of data management interfaces and underlying storing organization has adopted well-established medical approaches and was guided by expert neurosurgeons and neurologists. An important part of the system is a web-based environment for viewing and handling crucial electrophysiological recordings and neuro-imaging data. EEG analysis outcomes, measures and statistics are summarized in charts and, once calculated, are stored in the database for subsequent any-time access. EPILDA supports DICOM data management and provides automatic routines to extract and store header data such as patient, examination and scanner information. Imaging data acquired in various modalities can be loaded, and tools for image enhancement, filtering, edge detection and segmentation can be used to assist their interpretation. Efficient data management and analysis tools provided by EPILDA system, enable the evaluation of therapy in epilepsy, since valuable conclusions for the treatment outcome can be extracted. In addition to this, the results of patient's data analysis (both electrophysiological and neuro-imaging) can be used for the surgical candidacy determination in epilepsy. Innovative aspects and advantages:

  • Web-based, lightweight working environment, supported by any Java-enabled browser. It can be accessed over a secure connection by any network connected computer, given that valid credentials are provided.
  • Enhanced security policies to prevent from unauthorized access, ensure data integrity and fortify safe network functioning.
  • Patient-oriented browsing, efficient and ergonomic data organization design that minimizes "clicks per task".
  • Architectural design based on a central operator that handles peripheral attached mechanisms, adding flexibility, extensibility, reusability and maintainability.
  • Efficient medical image browsing, easy inspection and information-revealing filtering and analysis tools.
  • EEG analysis tools for the extraction of quantitative measures that are calculated once and stored in the database for subsequent any-time accessing.