PDA

View Full Version : Summary of Replies: (Bluetooth) IMU Sensor Studies



kvernest19
04-12-2009, 10:16 PM
Summary of Replies

Origional Question:

I have been asked to assist a physician at my institution with some
data collection. We are planning on using a Memsense Bluetooth IMU
sensor (http://www.memsense.com/images/downloads/81/Datasheet-(Rev.-D---Eff-Jan.-1,-2009).pdf)
to collect biomechanical forces on the body during exercise.
We are just wondering if:
1. Anyone else has used this device for Biomechanical Data collection.
2. Anyone has used a similar device for Biomechanical Data collection.
3. Any clinical studies have been done with this device in humans
4. Any other relevant experiences pertaining to the accuracy of the device
We have found some studies looking at downhill skiing and running, but
we were hoping we could use it to collect accelerations and rotations
during exercise, rather then things such as stride length and
location. Any input is appreciated and a summary of replies will be
posted.
Thanks,
--
Kyle Vernest
--
MHSc Clinical Biomedical Engineering Candidate
Institute of Biomaterials and Biomedical Engineering
University of Toronto, Canada




Replies:
-------------------------------------------------------------------------------------------------------------------------------------


We have the Sparkfun IMU v4. It has accelerometers, gyroscopes, and
magnetometers. I know there are plenty of accelerometers being used
in the literature, but I don't know of anyone making a full use of an
IMU. I hope that helps!
-Hyun Gu

Hyun Gu Kang, PhD

Research Assistant Scientist
Institute for Aging Research, Hebrew Seniorlife

--------------------------------------------------------------------------------------------------------------------------

I suspect you are aware of the FAB product (attached) and confess my
conflict of interest (we sell it). With that in mind, please review the
attached information as well as:
Performance Videos:
www.noraxon.de/biosyn/introduction.wmv - 4 MB stream
www.noraxon.de/biosyn/recording&playback.wmv - 16 MB stream
www.noraxon.de/biosyn/sensor&placement.wmv - 9 MB stream
www.noraxon.de/biosyn/demo biosyn hi res.zip - 43 MB download archive in
case the video stream is not running
Info-Material for the Biosyn FAB System:
Biosyn/FAB_InverseDynamicalTechniques.pdf - Background infos for the
inverse dynamics techniques used in Biosyn
www.noraxon.de/biosyn/fabsensor_range_accuracy_datarate_pdf - Data and Infos
related to sensor technology and accuracy
www.noraxon.de/biosyn/zhou2007 review motion tracking.pdf - Nice overview
article for motion capture systems


Best Regards,
Todd
Todd Shewman BA Kin.
Noraxon USA, Inc.
702-281-7903 (cell)
www.noraxon.com

--------------------------------------------------------------------------------------------------------------------------

You will need to be confident with transforming data from the local
sensor frame to the limb segment frame and the global reference frame
to make sense of the data. In addition, the heading vector provided by
the magnetometer and accelerometers in a static calibration will
drift. Even using a single IMU, the extraction of reliable, accurate
data is not a trivial exercise if there are no quasi-static phases
where the heading and g vectors may be re-established. In addition,
bluetooth will lose data if the line-of-sight connection is lost
unless you implement a robust data buffering scheme in the sensor. The
situation gets worse if you have fast changes of direction.

I suggest you take a look at the following papers:
Brodie, M., Walmsley, A., Page W., 2008 3D Dynamic Accuracy of
Inertial Measurement Units During Simple Pendulum Motion. Computer
Methods in Biomechanics and Biomedical Engineering, 11(03), pp. 235 -
242.
Brodie, M., Walmsley, A., Page W., 2008 Static Accuracy and
Calibration of Inertial Measurement Units (IMUs) for 3D Orientation.
Computer Methods in Biomechanics and Biomedical Engineering, 11(06),
pp. 641 - 648.
And
Brodie, M., Walmsley, A., Page, W., 2008 Fusion Motion Capture: A
Prototype System Using IMUs and GPS for the Biomechanical Analysis of
Alpine Ski Racing, Journal of Sports Technology, 1, 17-28
I would also suggest that you look around at the various IMUs
available before making a decision to purchase a single stand-alone
sensor.


--------------------------------------------------------------------------------------------------------------------------

> We are just wondering if:
> 1. Anyone else has used this device for Biomechanical Data collection.

We have not used the MemSense device. However it appears promising
because it has temperature compensation. It is fast enough for a single
sensor (150 Hz) but may have problems if you try to use more than one
IMU for multiple body segments. Bluetooth only supports seven devices
in a picoNet.

> 2. Anyone has used a similar device for Biomechanical Data collection.

We have built numerous custom IMUs and have used the WiTilt v3.0
http://www.sparkfun.com/commerce/product_info.php?products_id=8563
It is Bluetooth and simpler if you only need triaxial acceleration
and one gyro, and it has adjustable sampling rate.

We are currently building planar IMUs (biaxial
accelerometers and one gyro) with Zigbee radios that
allow 10 IMUs to be sampled synchronously at
50 Hz each.

We have also built very small Bluetooth triaxial
accelerometers that may be mass produced by a local
firm in the near future.

> 3. Any clinical studies have been done with this device in humans

Our planar IMUs will be used for sagittal analysis of occupational lifting.
The small Bluetooth triaxial accelerometers will be used for
clinical sleep research.

> 4. Any other relevant experiences pertaining to the accuracy of the device

We have been using homemade planar IMUs for over five years in
vehicle stability research. One key issue is to compensate for
thermal drift in zero offset and sensitivity of both MEMS accelerometers
and gyros. Secondly, magnetometers are necessary for full
3D work but can have problems near ferromagnetic objects.

HJSIII

H.J. Sommer III, Ph.D., Professor of Mechanical Engineering
Professor-In-Charge of Graduate Programs
Department of Mechanical and Nuclear Engineering
The Pennsylvania State University




---------------------------------------------------------------------------------------------------------------------------

Because the investigation of this new technology may be of interest to
everyone I have posted my response to Kyle's question about IMUs and
motion capture to the list.
It is quite possible to get accurate biomechanical data using such a
system including accelerations and rotations and then you can derive
net joint torques and forces. Theoretically IMUs should be more
accurate than video systems for this because the internal sensing
components measure rotation and acceleration directly, however there
are still a few pit falls to the new technology as there is with any
new technology. I have attached a list of my papers on the subject
below; information includes the problems I came across and how the
problems were overcome, and soon my PhD thesis will also be in the
public domain.
For a more graphic demonstration see my youtube videos at:
www.youtube.com/BrodieMAD For a short summary of how Fusion Motion
Capture works see the video titled: 'Science of Gold Medals and
Skiing', this work contributed to me being named Overall Runner Up,
2008 MacDiarmid Young Scientist of the year.

Regards,
Matthew Brodie

Publication list follows:
Brodie, M., Walmsley, A., Page, W., 2008 Fusion Motion Capture: Can
Technology be used to Optimise Ski Racing Technique? The Impact of
Technology on Sport 2, Page 825-831, 2008 Tailor & Francis Group,
London, ISBN-13: 978-0-415-45695-1
Brodie, M., Walmsley, A., Page, W., 2008 Fusion Motion Capture: A
Prototype System Using IMUs and GPS for the Biomechanical Analysis of
Alpine Ski Racing, Journal of Sports Technology, 1, 17-28
Brodie, M., Walmsley, A., Page W., 2008 3D Dynamic Accuracy of
Inertial Measurement Units During Simple Pendulum Motion. Computer
Methods in Biomechanics and Biomedical Engineering, 11(03), pp. 235 -
242.
Brodie, M., Walmsley, A., Page W., 2008 Static Accuracy and
Calibration of Inertial Measurement Units (IMUs) for 3D Orientation.
Computer Methods in Biomechanics and Biomedical Engineering, 11(06),
pp. 641 - 648.
Brodie, M., Walmsley, A., Page W., 2008 A comparison of motion
measurement using fusion motion capture and video motion analysis.
Computer Methods in Biomechanics and Biomedical Engineering, Under
Review
Brodie, M., Walmsley, A., & Page, W., 2008 How to ski faster: art or
science? In E. Müller (Ed.), Science and Skiing 4 (pp. 162-174).
Oxford: Meyer and Meyer Sport.