AN ULTRA-LOW POWER SUBBAND-BASED ELECTRONIC STETHOSCOPE
Julie Johnson
1
, David Hermann
1
, Melody Witter
1
, Etienne Cornu
1
, Robert Brennan
1
and Alain Dufaux
2
1
AMIS Canada, 611 Kumpf Drive, Unit 200, Waterloo, Ontario, Canada N2V 1K8
2
AMIS Switzerland, Champs-Montants 12a, 2074 Marin, Switzerland
ABSTRACT
Electronic stethoscopes are able to offer signal
amplification and other benefits over traditional
stethoscopes. However, many electronic stethoscopes rely
on a personal computer for their signal processing, which
reduces portability and requires a relatively large amount of
power. This paper presents a low power, portable electronic
stethoscope system that is based on a signal processing
approach using an over-sampled filterbank. This system is
implemented on an ultra-low resource DSP system. The
stethoscope incorporates multiple filtering modes as well as
audio record and playback (full and half speed)
functionality.
1. INTRODUCTION
Auscultation continues to be a routine part of
cardiovascular and pulmonary examinations. Recently,
there have been a number of advances in auscultation
technology, including the introduction of a number of
electronic stethoscopes [1,2,3]. Many of these stethoscopes
rely on the use of a personal computer to filter, record and
replay the signal. As such, these systems are large and have
limited portability. Little emphasis has been placed on
generating high quality audio in a small, portable unit.
Clearly, an electronic stethoscope system with the features
of these larger systems that is portable and low power
would be advantageous.
An electronic stethoscope system has been
implemented on a highly flexible ultra-low power,
miniature DSP system. At the heart of this system is an
over-sampled, weighted overlap-add (WOLA) filterbank
specifically designed for subband signal processing.
Multiple filtering modes, specialized for cardiovascular and
pulmonary auscultation applications, have been designed.
For recording, regular speed playback and half speed
playback, a low resource audio codec was developed
specifically for low frequency signals. Sufficient resources
remain for additional algorithms including heartbeat
monitoring and wireless functionality.
One of the most challenging aspects of the design is the
integration of algorithms requiring different filterbanks. A
method of fulfilling the requirements of two filterbanks
while minimizing computations and thus power
requirements, is presented. Filterbank selection is further
complicated by the low frequency nature of the input
signals. A method of combining the two required filterbanks
while effectively addressing this issue will be presented.
This method also accommodates the application of large
gain adjustments to decoded signals and is an essential
component of the half speed playback algorithm.
In the following sections, a description of the DSP
system architecture is presented, followed by an overview
of the stethoscope system. In Section 4, the implementation
of the stethoscope functionality is discussed in greater detail
showing innovative solutions to the design challenges
discussed above. Section 5 provides a characterization of
the stethoscope’s performance and conclusions are
presented in Section 6.
2. DSP SYSTEM
The DSP system consists of three major components: an
18-bit block floating point weighted overlap-add (WOLA)
filterbank coprocessor, a 16-bit fixed-point DSP core, and
an input-output processor (IOP). These components run in
parallel and communicate through interrupts and shared
memory. The parallel operation of these components
enables the implementation of complex signal processing
algorithms with low system clock rates and low resource
usage and is particularly adept at subband signal processing.
The WOLA coprocessor implements a flexible over-
sampled filterbank. Although initially designed for analysis
and synthesis involving over-sampled, complex subband
signals, it may be adapted to generate critically-sampled,
real-valued filterbanks as required for the codec in this
application [4].
The algorithms are implemented on the DSP system
using a 16-band, 4-times over-sampled WOLA filterbank
configuration with odd-stacking. The selected configuration
generates a group delay of 17 ms, consumes 4.1 mW, has a
system clock frequency of 5.12 MHz and a sampling
frequency (F
s
) of 8 kHz.
Copyright 2006 IEEE. Published in the 2006 International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2006),
scheduled for May 14-19, 2006 in Toulouse, France. Personal use of this material is permitted. However, permission to reprint/republish this
material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse
any copyrighted component of this work in other works, must be obtained from the IEEE. Contact: Manager, Copyrights and Permissions /
IEEE Service Center / 445 Hoes Lane / P.O. Box 1331 / Piscataway, NJ 08855-1331, USA. Telephone: + Intl. 908-562-3966.
