Pages

Monday, September 27, 2010

CS: The EPFL Real-Time Compressed Sensing-Based Personal Electrocardiogram Monitoring System

I previously mentioned the presentation of  Pierre Vandergheynst on the EPFL Compressive Sensing ECG system (check also the Q&A on the matter).. Pierre sent me more in the form of a webpage featuring the project A Real-Time Compressed Sensing-Based Personal Electrocardiogram Monitoring System by Hossein Mamaghanian, Karim Kanoun, and  Nadia Khaled 

From the website:

OUR APPROACH

Capitalizing on the sparsity of ECG signals, we propose to apply the emerging compressed sensing (CS) approach for a low-complexity, real-time and energy-aware ECG signal compression on WBSN motes. This is motivated by the observation that this new signal acquisition/compression paradigm is particularly well suited for low-power implementations because it dramatically reduces the need for resource- (both processing and storage) intensive DSP operations on the encoder side. The price to pay for these advantages however, is a more complex decoder. In general, reconstruction algorithms for CS recovery are computationally expensive, because they involve large and dense matrix operation, which prevents their real-time implementation on embedded platforms.
This project has developed a real time, energy-aware wireless ECG monitoring system. The main contributions of this work are: (1) a novel CS approach that precludes large and dense matrix operations both at compression and recovery, and (2) several platform-dependent optimizations and parallelization techniques to achieve real-time CS recovery. To the best of our knowledge, this work is the first to demonstrate CS as an advantageous real-time and energyefficient ECG compression technique, with a computationally light and energy-efficient ECG encoder on the state-of-the-art ShimmerTM wearable sensor node and a real-time decoder running on an iPhone as a representative of compact embedded (CE) device (acting as a WBSN coordinator).



This information has been added to the Compressive Sensing Hardware page.

No comments:

Post a Comment