Energy-efficient acquisition and embedded processing of bio-signals



Description

One of the greatest challenges in the design of modern embedded devices is energy efficiency and memory footprint. While the highly efficient microcontrollers and transmission devices have received a lot of attention from the industry and academia, sensor data acquisition and processing in medical devices is still based on a conservative paradigm that requires regular sampling at a high rate, in the order of hundreds or even thousands of times per second. Thus, smart algorithms and technologies are required to optimize the energy consumption and memory footprint while assuring data quality and integrity.

We work on developing novel biosignal sampling strategies that exploit the temporal properties and the expert knowledge we have to reduce the amount of acquired data by orders of magnitude. For this, we adopt the so-called event-based paradigm, in which the signal is only acquired and processed when something relevant is observed.

Following the event-based strategy it is possible to go even further by adapting the signal quality to the target task. For example, in the image above, we can see an electrocardiographic signal that is sampled with a traditional approach at 360 Hz. Below, the same signal is sampled with an adaptive event-based approach that takes into account the heart rate and the presence of physiological abnormalities. In this way, we are able to reduce the sampling rate to around 6Hz for most of the time, and increase it only when something unexpected is observed in the signal.

Related Publications

SPARE: A Spectral Peak Recovery Algorithm for PPG Signals Pulsewave Reconstruction in Multimodal Wearable Devices
Masinelli, Giulio; Dell'Agnola, Fabio Isidoro Tiberio; Arza Valdes, Adriana; Atienza Alonso, David
2021-04-13Conference PaperPublication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)Publication funded by SNF ML-edge (ML-edge: Enabling Machine-Learning-Based Health Monitoring in Edge Sensors via Architectural Customization)
An Event-Based System for Low-Power ECG QRS Complex Detection
2020-03-09DATE 2020 - Design, Automation & Test In Europe Conference & ExhibitionPublication funded by Human Brain Project  ()Publication funded by DeepHealth H2020 (Deep-Learning and HPC to Boost Biomedical Applications for Health)Publication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)
Event-Triggered Sensing for High-Quality and Low-Power Cardiovascular Monitoring Systems
Surrel, Grégoire ; Teijeiro, Tomas ; Chevrier, Matthieu ; Aminifar, Amir ; Atienza Alonso, David
2020IEEE Design & TestPublication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)Publication funded by Human Brain Project  ()
REWARD: Design, Optimization, and Evaluation of a Real-Time Relative-Energy Wearable R-Peak Detection Algorithm *
Orlandic, Lara; De Giovanni, Elisabetta; Arza, Adriana; Yazdani, Sasan; Vesin, Jean-Marc; Atienza, David
2019-10-072019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)Publication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)
e-Glass: A Wearable System for Real-Time Detection of Epileptic Seizures
D. Sopic, A. Aminifar, D. Atienza
2018International Symposium on Circuits and Systems (ISCAS)Publication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)
Self-Aware Wearable Systems in Epileptic Seizure Detection
F. Forooghifar, A. Aminifar, and D. Atienza Alonso
2018Euromicro Conference on Digital System Design (DSD)Publication funded by Hasler MyPreHealth (Predicting Episodic Disorders with Health Companions)
A wearable device for physical and emotional health monitoring
S. Murali, F. Rincon, D. Atienza
2015Computing in Cardiology Conference (CinC)Publication funded by Nano-Tera ()