Sensor Fusion and Resource Management for Physiological Monitoring in the Swarm
Viswam Nathan, Ilge Akkaya, Roozbeh Jafari

Citation
Viswam Nathan, Ilge Akkaya, Roozbeh Jafari. "Sensor Fusion and Resource Management for Physiological Monitoring in the Swarm". Talk or presentation, 14, October, 2015; Poster presented at the 2015 TerraSwarm Annual Meeting.

Abstract
Ubiquitous and pervasive physiological monitoring in the context of the swarm incurs the obstacle of potentially very noisy signals, for example caused by motion artifacts on wearable sensors. The system would also need to recognize when one or more sensors is not being used or improperly placed so that the estimate can come from any remaining sensors. Finally, ensuring efficient usage of power and computational resources in a system of multiple sensors on the body and in the environment is of primary importance. We use a particle filter to ensure robust estimation of target phenomenon in noisy scenarios. This probabilistic technique has several advantages, among which are the fact that it does not require linear models of signal and noise to operate and also the fact that it can track multiple possible states of the system as it converges on a posterior probability distribution of the true state. We established the efficacy of the filter by showing that it can estimate the true heart rate in real motion affected electrocardiogram (ECG) data with an error of less than 5 beats per minute (5bpm) under SNR conditions as low as -6dB. We then present ongoing research that includes models for how the observations from multiple sensors can be fused to improve the accuracy, as well as an information theoretic approach to dynamically minimize the number of sensors being used to save on resources while maintaining estimation accuracy. We are testing these models on a real multi-modal dataset (ECG + Photoplethysmogram + Accelerometer) from the Signal Processing Cup 2015 and the initial results are promising in terms of improved estimation accuracy. This framework can prove very useful to ensure effective exploitation of multiple sensors and modalities to counteract noisy signals, while also maintaining awareness of resource usage to facilitate extended use of the sensors for long term monitoring.

Electronic downloads


Internal. This publication has been marked by the author for TerraSwarm-only distribution, so electronic downloads are not available without logging in.
Citation formats  
  • HTML
    Viswam Nathan, Ilge Akkaya, Roozbeh Jafari. <a
    href="http://www.terraswarm.org/pubs/657.html"><i>Sensor
    Fusion and Resource Management for Physiological Monitoring
    in the Swarm</i></a>, Talk or presentation,  14,
    October, 2015; Poster presented at the <a
    href="http://terraswarm.org/conferences/15/annual"
    >2015 TerraSwarm Annual Meeting</a>.
  • Plain text
    Viswam Nathan, Ilge Akkaya, Roozbeh Jafari. "Sensor
    Fusion and Resource Management for Physiological Monitoring
    in the Swarm". Talk or presentation,  14, October,
    2015; Poster presented at the <a
    href="http://terraswarm.org/conferences/15/annual"
    >2015 TerraSwarm Annual Meeting</a>.
  • BibTeX
    @presentation{NathanAkkayaJafari15_SensorFusionResourceManagementForPhysiologicalMonitoring,
        author = {Viswam Nathan and Ilge Akkaya and Roozbeh Jafari},
        title = {Sensor Fusion and Resource Management for
                  Physiological Monitoring in the Swarm},
        day = {14},
        month = {October},
        year = {2015},
        note = {Poster presented at the <a
                  href="http://terraswarm.org/conferences/15/annual"
                  >2015 TerraSwarm Annual Meeting</a>},
        abstract = {Ubiquitous and pervasive physiological monitoring
                  in the context of the swarm incurs the obstacle of
                  potentially very noisy signals, for example caused
                  by motion artifacts on wearable sensors. The
                  system would also need to recognize when one or
                  more sensors is not being used or improperly
                  placed so that the estimate can come from any
                  remaining sensors. Finally, ensuring efficient
                  usage of power and computational resources in a
                  system of multiple sensors on the body and in the
                  environment is of primary importance. We use a
                  particle filter to ensure robust estimation of
                  target phenomenon in noisy scenarios. This
                  probabilistic technique has several advantages,
                  among which are the fact that it does not require
                  linear models of signal and noise to operate and
                  also the fact that it can track multiple possible
                  states of the system as it converges on a
                  posterior probability distribution of the true
                  state. We established the efficacy of the filter
                  by showing that it can estimate the true heart
                  rate in real motion affected electrocardiogram
                  (ECG) data with an error of less than 5 beats per
                  minute (5bpm) under SNR conditions as low as -6dB.
                  We then present ongoing research that includes
                  models for how the observations from multiple
                  sensors can be fused to improve the accuracy, as
                  well as an information theoretic approach to
                  dynamically minimize the number of sensors being
                  used to save on resources while maintaining
                  estimation accuracy. We are testing these models
                  on a real multi-modal dataset (ECG +
                  Photoplethysmogram + Accelerometer) from the
                  Signal Processing Cup 2015 and the initial results
                  are promising in terms of improved estimation
                  accuracy. This framework can prove very useful to
                  ensure effective exploitation of multiple sensors
                  and modalities to counteract noisy signals, while
                  also maintaining awareness of resource usage to
                  facilitate extended use of the sensors for long
                  term monitoring.},
        URL = {http://terraswarm.org/pubs/657.html}
    }
    

Posted by Viswam Nathan on 10 Oct 2015.
Groups: services

Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright.