Visualization of Multivariate Physiological Data for Cardiorespiratory Fitness Assessment through ECG (R-Peak) Analysis
03/08/2016 - UPDATE: You can now download PhysioLab from here: http://neurorehabilitation.m-iti.org/tools/physiolab
The recent rise and popularization of wearable and ubiquitous fitness sensors has increased our ability to generate large amounts of multivariate data for cardiorespiratory fitness (CRF) assessment. Consequently, there is a need to find new methods to visualize and interpret CRF data without overwhelming users. Current visualizations of CRF data are mainly tabular or in the form of stacked univariate plots. Moreover, normative data differs significantly between gender, age and activity, making data interpretation yet more challenging.
For that reason we developed the PhysioLab, a novel CRF assessment tool based on physiological computing and radar plots that provides a way to represent multivariate cardiorespiratory data from electrocardiographic (ECG) signals within its normative context. To that end, 5 parameters are extracted from raw ECG data using R-peak information: mean HR, SDNN, RMSSD, HRVI and the maximal oxygen uptake, VO2max. Our tool processes ECG data and produces a visualization of the data in a way that it is easy to compare between the performance of the user and normative data. This type of representation can assist both health professionals and non-expert users in the interpretation of CRF data.
PhysioLab uses different signal pre-processing methods
to reduce low frequency artifacts caused by respiration and
user movements. To stabilize ECG baseline wander, which is
characterized for baseline oscillations at a very-low frequency
drifting between 0.15 Hz and 0.3 Hz, some denoising
techniques are used. First, the signal is smoothed by a fifthorder
Savitzky-Golay FIR filter. Then, a low order
polynomial is fitted to the raw data and is used to detrend the
ECG signal. At this stage, a simple threshold rule based on
the peak morphology is applied to detect the R-peaks. All ECG parameters are displayed within the user interface of
the PhysioLab toolbox and stored to produce a Radar Plot, consisting of performance profiles derived from available normative
data are generated and color-coded facilitating the
interpretation and comparison of user performance (blue line)
with the CRF levels (high-green, average-yellow, low-red).
For more information, please refer to the following paper:
----------------
The recent rise and popularization of wearable and ubiquitous fitness sensors has increased our ability to generate large amounts of multivariate data for cardiorespiratory fitness (CRF) assessment. Consequently, there is a need to find new methods to visualize and interpret CRF data without overwhelming users. Current visualizations of CRF data are mainly tabular or in the form of stacked univariate plots. Moreover, normative data differs significantly between gender, age and activity, making data interpretation yet more challenging.
For that reason we developed the PhysioLab, a novel CRF assessment tool based on physiological computing and radar plots that provides a way to represent multivariate cardiorespiratory data from electrocardiographic (ECG) signals within its normative context. To that end, 5 parameters are extracted from raw ECG data using R-peak information: mean HR, SDNN, RMSSD, HRVI and the maximal oxygen uptake, VO2max. Our tool processes ECG data and produces a visualization of the data in a way that it is easy to compare between the performance of the user and normative data. This type of representation can assist both health professionals and non-expert users in the interpretation of CRF data.
PhysioLab uses different signal pre-processing methods
to reduce low frequency artifacts caused by respiration and
user movements. To stabilize ECG baseline wander, which is
characterized for baseline oscillations at a very-low frequency
drifting between 0.15 Hz and 0.3 Hz, some denoising
techniques are used. First, the signal is smoothed by a fifthorder
Savitzky-Golay FIR filter. Then, a low order
polynomial is fitted to the raw data and is used to detrend the
ECG signal. At this stage, a simple threshold rule based on
the peak morphology is applied to detect the R-peaks. All ECG parameters are displayed within the user interface of
the PhysioLab toolbox and stored to produce a Radar Plot, consisting of performance profiles derived from available normative
data are generated and color-coded facilitating the
interpretation and comparison of user performance (blue line)
with the CRF levels (high-green, average-yellow, low-red).For more information, please refer to the following paper:
Muñoz, J., Bermúdez i Badia, S., Cameirão, M., & Gouveia, E. (2015). Visualization of Multivariate Physiological Data for Cardiorespiratory Fitness Assessment through ECG (R-Peak) Analysis. In 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE. (Download) (Cite)
Comments