Measuring Caloric Intake Using Chewing Sounds

1
Measuring Caloric Intake Using Chewing Sounds
Jimmy Fong (Co-Leader), Matt Valaskey
(Co-Leader), Aditi Bharatkumar (Communications),
Bryan Mounce (BWIG), Vidhya Raju (BSAC)
The goal of the project is to create a device
capable of monitoring chewing sounds. Analysis
of this data will recognize types of foods eaten.

• In-Ear Circuit Design
• Combination of proposed design ideas
• - based on custom amplifier circuit design
• - is the most cost-effective
• - requires microphone to be inserted in ear
  canal
• - closest to highest region of bone conduction
  (maximal exposure to chewing sounds)
• - optimal insulation from ambient noise

• In-ear probe microphone design
• - places microphone in ear canal
• External microphone design
• - mastoid process
• - throat
• - behind ear
• Custom Amplifier Circuit Design

Audioscan Professional Audiology Microphone and
Amplifier
Amplification with a Custom Circuit Board
Spectral Analysis in Matlab
FFT Analysis in Excel
Final Circuit Board Amplifier and Microphone
Earpiece

• Collection of preliminary chewing sound data from
  different microphone locations
• Analysis and evaluation of signal quality based
  on microphone location in-ear microphone optimal
• Recording of chewing data
• Analysis of collected data using Excel and
  Matlab Excel analysis superior
• Construction of an economical circuit-board design

The acoustic properties of chewing as measured in
the ear canal

• The microphone should
• Have a frequency response of 20 Hz to 20000 Hz
  to ensure that all chewing sounds are recorded.
  
• Reproducibly record chewing.
• Analysis of Data should
• Distinguish between chewing frequencies and
  normal speech/ambient noise frequencies
• Possibly reveal a correlation between chewing
  frequency and caloric intake
• Ergonomics
• Device should fit comfortably in the ear and
  should be minimally intrusive
• Hygiene should not be compromised probe tubes
  for the microphone should be disposable
• Cost should not over exceed the benefit of the
  device.

Further investigation will include further
testing of the correlation between peak frequency
and type of food consumed. In addition, a
portable version of the device will be developed.
The in-ear circuit design will aid in tracking
dietary intake with the greatest accuracy.
Manufacture of First Microphone
Early Circuit Board Construction
National Institute of Health Statistics.
http//win.niddk.nih.gov/statistics/index.htmwhyd
odiffer Amft, O., Stager, M., Lukowicz, P., and
Troster, G. 2005. Analysis of chewing sounds
for dietary monitoring. Lecture Notes in
Computer Science 56-72. Cichero, J. A. and
Murdoch, B. 2002. Acoustic signature of the
normal swallow characterization by age, gender,
and bolus volume. Ann Oral Rhinol Laryngol 111
623-631. Doctronics Education Publishing. 2006.
http//www.doctronics.co.uk/scope.htmaudio Maxim
Integrated Products. 2006 Crunching FFTs with
Microsoft Excel. http//www.maxim- ic.com/appnotes
.cfm/an_pk/3292
Microphone Sounds Displayed on an Oscilloscope
Initial Computer and Technical Analysis