Title: STARR – Decision Support and Self-Management System for Stroke Survivors
Funding source: H2020 PHC-28-2015
Partners: CEA (coordinator), University of Lund, Osakidetza, The Stroke Association, Fondation Hopale, Fiz Karlsruhe, Blulinea, RT-RK, Telefonica
Principal investigator: Dr. Djamila Aouada
Researchers: Dr. Enjie Ghorbel, Dr. Abd El Rahman Shabayek, Renato Baptista
Starting date/ Duration: 01/02/2016 – 42 months

Stroke a leading cause of death and disability, with an estimated total cost of €65 billion per year in Europe. Even though preventive measures are in place to reduce the incidence of stroke, the number of persons having a stroke in Europe is likely to increase from 1.1 million/year in 2000 to more than 1.5 million/year in 2025 because of the increasing ageing population. Secondary stroke carries with it a greater risk than first-ever stroke for death and disability. Also, as mortality from first strokes has decreased recently, the number of people at risk for a secondary stroke has increased, with an associated increase in healthcare costs. In order to reduce these stroke statistics and the associated cost, the self-management of stroke risk factors is particularly suitable and necessary for the following reasons: 1) risk factors for stroke are well-known, and 2) 90% of strokes or secondary stroke events are preventable if the risk factors are managed appropriately. The Decision SupporT and self-mAnagement system for stRoke survivoRs (STARR) project and the system developed in it are targeting the self-management of stroke risk factors. Based on existing computational predictive models of stroke risk, we will develop a modular, affordable, and easy-to-use system, which will inform stroke survivors about the relation between their daily activities (e.g., medication intake, physical and cognitive exercises, diet, social contacts) and the risk of having a secondary stroke. This will lead to better prevention and a reduction of the number of secondary stroke events, as well as to a more efficient participation of patients in medical decision-making. A multidisciplinary consortium has been built for achieving the objectives of this ambitious project, involving stroke survivors’ associations, healthcare actors, sensing and human-machine interfaces experts. The consortium also comprises 3 European companies which will exploit the results of the project after its end.

Publications

Temporal 3D Human Pose Estimation for Action Recognition from Arbitrary Viewpoints
Adel Musallam, MohamedBaptista, RenatoAl Ismaeil, KassemAouada, Djamila
in 6th Annual Conf. on Computational Science & Computational Intelligence, Las Vegas 5-7 December 2019 (2019, December)

VIEW-INVARIANT ACTION RECOGNITION FROM RGB DATA VIA 3D POSE ESTIMATION
Baptista, RenatoGhorbel, EnjiePapadopoulos, KonstantinosDemisse, GirumAouada, DjamilaOttersten, Björn
in IEEE International Conference on Acoustics, Speech and Signal Processing, Brighton, UK, 12–17 May 2019 (2019, May)

Home Self-Training: Visual Feedback for Assisting Physical Activity for Stroke Survivors
Baptista, RenatoGhorbel, EnjieShabayek, Abd El RahmanMoissenet, FlorentAouada, DjamilaDouchet, AliceAndré, MathildePager, JulienBouilland, Stéphane
in Computer Methods and Programs in Biomedicine (2019)

A View-invariant Framework for Fast Skeleton-based Action Recognition Using a Single RGB Camera
Ghorbel, EnjiePapadopoulos, KonstantinosBaptista, RenatoPathak, HimadriDemisse, GirumAouada, DjamilaOttersten, Björn
in 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, Prague, 25-27 February 2018 (2019, February)

Two-stage RGB-based Action Detection using Augmented 3D Poses
Papadopoulos, KonstantinosGhorbel, EnjieBaptista, RenatoAouada, DjamilaOttersten, Björn
in 18th International Conference on Computer Analysis of Images and Patterns SALERNO, 3-5 SEPTEMBER, 2019 (2019)

Anticipating Suspicious Actions using a Small Dataset of Action Templates
Baptista, RenatoAntunes, MichelAouada, DjamilaOttersten, Björn
in 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISAPP) (2018, January)

Deformation-Based Abnormal Motion Detection using 3D Skeletons
Baptista, RenatoDemisse, GirumAouada, DjamilaOttersten, Björn
in IEEE International Conference on Image Processing Theory, Tools and Applications (IPTA) (2018, November)

Key-Skeleton Based Feedback Tool for Assisting Physical Activity
Baptista, RenatoGhorbel, EnjieShabayek, Abd El RahmanAouada, DjamilaOttersten, Björn
in 2018 Zooming Innovation in Consumer Electronics International Conference (ZINC), 30-31 May 2018 (2018, May 31)

Flexible Feedback System for Posture Monitoring and Correction
Baptista, RenatoAntunes, MichelShabayek, Abd El RahmanAouada, DjamilaOttersten, Björn
in IEEE International Conference on Image Information Processing (ICIIP) (2017)

Video-Based Feedback for Assisting Physical Activity
Baptista, RenatoGoncalves Almeida Antunes, MichelAouada, DjamilaOttersten, Björn
in 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISAPP) (2017)

STARR – Decision SupporT and self-mAnagement system for stRoke survivoRs Vision based Rehabilitation System
Shabayek, Abd El RahmanBaptista, RenatoPapadopoulos, KonstantinosDemisse, GirumOyedotun, OyebadeAntunes, MichelAouada, DjamilaOttersten, BjörnAnastassova, MargaritaBoukallel, MehdiPanëels, SabrinaRandall, GaryAndré, MathildeDouchet, AliceBouilland, StéphaneOrtiz Fernandez, Leire
in European Project Space on Networks, Systems and Technologies (2017)

Visual and human-interpretable feedback for assisting physical activity
Goncalves Almeida Antunes, MichelBaptista, RenatoDemisse, GirumAouada, DjamilaOttersten, Björn
in European Conference on Computer Vision (ECCV) Workshop on Assistive Computer Vision and Robotics Amsterdam, (2016)

TEST

A dataset containing 400 real, high-resolution human scans of 200 subjects (100 males and 100 females in two poses each) with high-quality texture and their corresponding low-resolution meshes, with automatically computed ground-truth correspondences. See the following table.

TEST

A dataset containing 400 real, high-resolution human scans of 200 subjects (100 males and 100 females in two poses each) with high-quality texture and their corresponding low-resolution meshes, with automatically computed ground-truth correspondences. See the following table.

TEST

A dataset containing 400 real, high-resolution human scans of 200 subjects (100 males and 100 females in two poses each) with high-quality texture and their corresponding low-resolution meshes, with automatically computed ground-truth correspondences. See the following table.

TEST

A dataset containing 400 real, high-resolution human scans of 200 subjects (100 males and 100 females in two poses each) with high-quality texture and their corresponding low-resolution meshes, with automatically computed ground-truth correspondences. See the following table.