Biomedical Telemetry - A Review and Overview

Item Type text; Proceedings

Authors Caceres, C. A.; Slater, L. E.

Publisher International Foundation for Telemetering

Journal International Telemetering Conference Proceedings

Rights Copyright © International Foundation for Telemetering

Download date 09/10/2020 05:58:20

Link to Item http://hdl.handle.net/10150/607013

 BIOMEDICAL TELEMETRY-A REVIEW AND OVERVIEW*

C. A. CACERES, M.D.
Consultant
Washington, D.C.

L. E. SLATER
Technical Representative
Commonwealth of Puerto Rico.

Summary Biomedical telemetry has yet to fulfill the high promise generated by its
first significant use ten years ago. Most of the progress has been in research
biotelemetry; in the monitoring of animal physiology and behavior to gain new insight
into both the normal and pathological state. Many clinical applications have developed
but few, with the notable exceptions of telephone telemetry systems in cardiology and
the monitoring of astronauts, have achieved wide acceptance. The nub of the problem
has been the bias in equipment design and system orientation towards research criteria.
There is a compelling need and opportunity for telemetric systems specifically designed
for clinical, or on-the-spot, use.

Introduction Biomedical telemetry today stands on the threshold of widespread,

valuable use in clinical medicine and health care. Its worth as a technique to gain
continuous physiologic data which can be machine analyzed is proven. The quality and
unique nature of the information it can produce is universally acknowledged. All signs
are “go” for the harvest of this new technology.

Still, despite this progress and acknowledgement, biomedical telemetry seems to be held
back by some elusive barrier; a constraining force which keeps it from the great field of
clinical applications.

How far has biomedical telemetry really come? What are the problems that still inhibit it
as a proven clinical technique? Where is the technique going? These are the questions
we will try to answer in this brief review paper.

Let us start by considering the basic contribution that biomedical technology offers: Its

*
The deadline for submitting this paper prevented inclusion of results of a national survey to
determine recent progress in biomedical telemetry. This material will be presented at the
Conference and available in printed form afterwards from the authors.
ability to organize a research or clinical problem so that it is amenable to the “systems”
approach. All “systems” comprise an input, a medium to be traversed, and an output. In
medicine and biology, telemetry can be the integrating force which defines the problem
as a system. This is the crux of its contribution. Without this integrating force, many
medical instruments today fail to achieve their true value--they fail to create a system and
simply remain components from which full advantage is not gained.

While the role of biomedical telemetry is to organize information into a systems

approach, the goal of the technique is more specific: namely to faithfully convey
phenomena so that they are useable units of information. These information units can tie
into the normal thinking process of a human being or the physiological mechanisms of
the subject. They can similarly tie into a machine interface (for subsequent analysis by
the human being), or for action by the machine serving as a control system. Hence the
goal of biomedical telemetry--to gain useable information--is what makes possible its
role--the role of facilitating the systems approach.

Progress in Technique Considerable progress has taken place over the past ten years
in refining and improving the technical aspects of biomedical telemetry. A brief review
of this progress will help clarify the difficulties which seem to curb its wider clinical
application.

First let us consider the media being used. In the narrower technical sense, biomedical
telemetry today usually refers to special informational conversion or transmission
generally by radio or telephone. But this does not exclude combinations of the two or
lasers, ultra-sound, infrared, and other media which are not yet generally available to all
of us.

Whatever the mode of transmission chosen, coding of the data should be devised to save
space in the media, to convey not only signal but related information, and to allow for
feedback. The media should optimally be adaptive depending on the needs of the signal
at specific times.

A practical telemetric system design must include or interface with filtering for noise or
preemphasis of the signal, validity checks, buffering, longer-term storage, data reduction,
feedback, control circuits, and displays.

For good information transmission we have come to learn that different personnel are
required to assume responsibility for three specific areas in telemetry:

First there is the information source, including the transducer as well as the clinical
systems involved. Their characteristics and variability must be established as firmly as is
possible. The transducer’s message results from interaction between the time-varying
characteristics of a living source and its relatively invariant (or unadjusting) circuits and
components. It has been found necessary for both the engineer and biologist to
comprehend the transducer and information source equally well. This does not mean that
they each must specifically know the background rationale of why the characteristics
exist in transducer or biologic source. But, rather, that they must act as a team.

The second area of concern involves the amplification factor, noise, and the transmitter
which codes the message or signal. The methodology for this is within the realm of
today’s standard electrical engineering practice, and these items are open to creativity on
the part of the engineer. The biologist need not be more than passively knowledgeable
about this technique if he has conveyed to the engineer the true sense of what the
biologic system can be expected to do.

The third area includes the receiver and the subsystem responsible for decoding and final
transmission of the information to its user. Technical details here are primarily the
concern of the engineer.

Evaluation of final information the system produces is a matter for both biologist and
engineer. The display can mislead if how it is produced is not comprehended. The final
checkout requires that the interpretation makes sense, not just to the biologist but to the
engineer as well.

Noise and Distortion Superimposed on the telemetry system is a noise source which
has come to be considered an undesirable but always present element. Individual
component noise factors are significant, but in providing the user with performance
specifications only the total value is of importance. The more components, the more
interfaces, the longer the transmission distances--all of these tend to increase noise.
Higher-quality equipment may decrease it, but cost then enters the noise picture and
tradeoffs are necessary. Numerous noise reducing methods can be considered by the
engineer, but it is clear today that a good telemetric system need not be completely
noiseless. Optimization is what we are after for useable information units.

Practical applications of biomedical telemetry have proven we can cope with noise of
certain values, and we may even wish to have certain distortions to aid us in our analysis.
For example, high frequency-response loss in equipment results in slurring of the signal
which rounds off corners or peaks. Amplitudes are diminished and minute waves
disappear. But, this distortion may be useful in evaluating the electrocardiographic signal
from a subject with extraneous low-level random noise. Low frequency distortion is
helpful, as another example, in fetal electrocardiography. Maternal abdominal motion
and respiration can be eliminated for a better fetal electrocardiogram by elimination of
low frequencies, i.e. distortion. A goal of a good telemetric system is thus transmission
of known or established quality and characteristics.
Data Characteristics The combined effect of resistance, capacitance, and inductance
depends on their individual values in the system as well as on the nature of the input
signal. The engineer can calculate this. He has learned he should contrast hardness of
that type of data with medical, empiric soft data.

Biologic and medical signals have characteristics which are often due to historical as
well as physical or even physiological considerations. Biology was early studied with
devices with poor high-frequency response. String galvanometers and the like generally
had good low frequency response. Body signals, however, have been studied with these
for decades, and relationships between instruments and body characteristics were
established by trial and error. In the past we established dogma based on knowledge of
the basic signal, or the data acquisition system. In effect, in the past 100 years we
developed a cryptographic method of analysis of data generally recorded with
instrumentation of reasonable low-frequency response and without complete analysis of
input characteristics. Nevertheless clinicians have used the data and achieved reasonable
results.

Miniaturization Sensors or transducers should not change what the subject would
ordinarily engage in nor should they encumber an ill patient. Hence, miniature radio
transmitters attached to the sensor or transducer have been found to be a good way to
reasonably collect medical and biological data outside of a strict laboratory environment.

Miniaturized instrumentation for transducers and sensors was given a healthy stimulus
by the astronaut program. Endoradiosonds in biological experiments relating to animal
behavior, or for intracavitary organ data, for example, can transmit values of pressure,
temperature, acidity, oxygen tension, blood pressure, radiation intensity, motion,
bleeding sites, and other vital information such as the electrocardiogram and the
electroencephalogram. Too few researchers and clinicians have made use of these. In
part this has been because of cost.

With miniaturization, telemetry has been used not only to obtain data, but to provide a
signal stimulus. This and the feedback are significant contributions in the field of
research. Telemetric systems can of course trigger electrical, thermal, or any other type
of stimuli. The results suggest future methods for the study of human responses. Now
that the physiologist is able to transmit the stimulus and observe results in one’s own
environment, several types of studies are possible.

Today there are active and passive transmitters. The passive transmitter requires no
battery within its capsule and obtains its energy from external sources. The capability of
small transmitters as devices that trigger servo-mechanisms within the body must be
mentioned as a future possibility.
Telephone Transmission The most widely used medical telemetry system is the
telephone. Telephone transmission of medical signals has three principal justifications.

The first of these is the convenience due to its ubiquity. Telephone transmission systems
are able to eliminate certain problems involving identification of the patient,
intermediate storage, transportation to the central site. This may all sound too simple.
But, by sending a signal to its ultimate destination at the time it is acquired from the
patient, and thus eliminating intermediate manipulation, the path to on-line processing of
all data becomes cleared for mass use.

A second justification has been found to be time saving. When the distances involved are
long, telephone transmission can allow data to be sent and returned while it is still
usable. In an emergency situation, this advantage of telephone telemetry is obvious. This
particular value was one of the first to be widely recognized as consulting cardiologists
had telephone data receivers installed in their home and office. These allowed them to
give informed consultations immediately, without risk of life-threatening delay.

The third justification of telephone telemetry, one that is not immediately apparent, is its
“translation ability”. In order to receive signals or data from any patient location or
laboratory, a center need have only a telephone receiver. Whether the data is recorded on
various types and brands of tape recorders, or in different format, or with different
electrical specifications, it will enter the center via the telephone in a universal form and
be immediately adaptable to the center’s own equipment. The telephone supplies a
matching interface for transmission of data from machine to machine, regardless of
individual characteristics.

The first transmission of medical data by hardwire and components of a telephone

system was in 1903 by Einthoven who recorded electrocardiograms from a hospital and
transmitted them to his laboratory where his string galvanometer amplified the signal.

These pioneer efforts did not stimulate much other than the inclusion of hardwire in a
few hospitals for transmission of electrocardiograms from wards to the
electrocardiographic heart station. Part of the problem, of course, was the lack of
appropriate amplification and filtering, etc., that is lack of electronic telemetric systems.

Many of the problems were solved with the availability of FM systems and thier
decreased sensitivity to variations in signal amplitude as opposed to amplitude
modulation. This and the fact that the systems require less gain in the amplifiers and are
inherently less noisy than amplitude modulation led to the now almost universally
accepted usage of FM in medical telemetric systems.
The first commercially available medical telephone package appeared in 1958 designed
for electrocardiographic transmission. Although somewhat difficult to operate it was
used with some acceptance in the Kansas City area to allow practitioners to send
electrocardiograms from their offices to that of cardiologists for interpretation of the
signals.

Transmission of data by telephone is fairly well stabilized today. Three channel

multiplexing of medical signals is a routine in many cases. At the present time, we are
generally capable of reasonable quality transmission with existing telephone systems in
the United States and, to a large extent, elsewhere. The telephone companies maintain
certain tariffs for their data sets. Medical-data transmission from others’ data sets also fit
readily into their standard data systems.

Regretfully there is still lacking sufficient interplay between the telephone companies
and the manufacturer of the transducers and interface devices that can be used to send
data through the telephone. Often custom connections must be made. It would be
desirable to have standardization of output from every medical transducer in order to
have direct input into a telephone either by direct connection or audio transmission.
Higher quality would result. This area should be investigated by the manufacturers and
the telephone companies for their and, particularly, the users benefit. Medical
instrumentation groups should also look into this area of public interest.

Another area for suitable cooperation would be in the development of coding systems to
allow the user to transmit identifying subject codes and other ancillary data in a format
suitable for the telephone as well as the transducer. Identification mechanisms are an
indispensable item that has been left off most medical instrumentation in the past
producing huge record-room problems. Too much effort is now being put into solving
the record room problem without looking into the initial cause of the problem. These
initial coding problems are easily solvable by automation systems, if the coded
information is carried along with or preceding the signal. This is a job for the telemetric
system. The idea of voice communication for coding on the same channel is inadequate.

Radio Transmission The first radio transmission occurred in 1921 when the U.S.
Army SignaTCorp used a radio telemetry system for heart sounds. It was intended that
these be recorded on ships without physicians on board so that facilities on shore could
do the actual analysis.

With radio as regulated by the Federal Communications Commission, it should be noted

that a system that interferes with other radio operations can be considered illegally
operated, regardless of conformity to general specifications. The first medical radio
license was granted for medical signal transmission from ambulances. The definite or
special (if there be any) requirements for medical or biological radio transmission, either
short or long range, have not been established. In part this is due to lack of users and
expertise. Thus, rules and regulations standardizing equipment and modes of
transmission may yet be premature, other than as generally stated by the FCC for other
similar users.

A Model for Medical Telemetry Perhaps the most remarkable and finest model of
biomedical telemetry exists in the astronaut monitoring programs. Yet it is the sad truth,
both in reviewing telephone and radio, that we are years behind in adapting this to
medicine, as contrasted to space technology contributions to other fields. The astronaut
programs combine various methods of telemetric communication and a multitude of
advances in the field of electronic technology. The worldwide network of tracking
stations both on land and on ships allows inflight information to be transmitted and
received for immediate analysis.

Astronaut monitoring has allowed display on earth of a multiplicity of biological signals.

Temperature, electrocardiogram, respiratory curve, and blood pressure are of course the
most significant for vital sign determination. Electroencephalograms and
electromyograms are other suitable signals.

In general the number and type of signals required for space monitoring are no different
than the class of signal required by the anesthetist in the operating room, the physician in
his intensive care suite, or the biologist in the research laboratory. We must match the
advances made by the astronaut program with the advances made in the medical world in
interpretation of the signals by computers and have a real-time on-line aid to the medical
monitor.

There is also no difference in the volume and presentation of data required by the
clinician or the space-flight monitor. Trend information has been shown to be more
reliable than single values. This fact is known in medical practice but is not used as
extensively as it should be. Trend data today are usually obtainable after the fact.

However, large masses of data can have immediate analysis by statistical techniques.
Computers can handle the data and provide indices of where a subject is in reference to
previous training experience or where the subject is in terms of others of similar
characteristics or disease.

Clinical Applications Requirements placed on clinical systems of biomedical

technology are that components must have reasonable sensitivity, stability, and reliable
performance as required for the specific clinical work.

A typical application is to detect the energy from the fetal heart. The utility of radio
telemetry in this situation is evident. To initially detect the signal, wires need not be
strung on the mother’s abdomen. Wires add artifacts due to movement and induction
from AC power lines. Additionally, the subject is more comfortable without them.

Urbach developed a practical system for delivery suites which allows one to monitor the
birth process of several subjects. He first recorded from the abdomen of the pregnant
female. After the fetal membranes rupture, he attached electrodes to the scalp of the child
during birth. To these electrodes are attached miniature transmitters. Neither the birth
process nor the mother is disturbed and constant monitoring is possible. The delivery
suite can be wired for antennae placed in all required locations in several rooms and
hallways at different positions so as not to miss any of the signals. Once received at the
antenna the signals are transmitted to a central control room. The whole procedure does
not contaminate the area or obstruct it with wires but does afford proper observation.
Transmission is necessary for only 50-100 feet. This might well be the practical
requirement in most hospital situations that can be envisioned.

Perhaps the greatest potential in the near future for radio telemetry in a hospital situation
will be in intensive care suites required to monitor patients with heart attacks or
postsurgical suites or the operating room itself. In the distant future the principles
learned from these can be incorporated into monitors for every patient in the hospital.
This would diminish the requirements for increasing nursing personnel and could
diminish their constant performance of basic necessary but routine monitoring activities
that do not require the skill that the nurse possesses. Hospital design will be markedly
changed as a result of these advances.

Pioneering systems of biomedical telemetry have been described for neurosurgical

procedures, anesthetic, surgical, and other specialized needs. The primary problem is
determination of what parameters are actually needed, how much data is required from
these parameters and data presentation modes in each specific circumstance.

Telemetric monitoring offers an advantage noted particularly in coronary care units.

Hazardous conditions, such as ventricular tachycardia,detected early, can in most
instances be successfully treated without fatality. Monitoring systems to sound an alarm,
offer remote control, and initiate therapeutic measures require the use of on-line
computers to analyze and interpret the data. Telephone telemetry of the initially radio
received signal incorporated into a computer system (i.e., integration of various
subsystems) offers an important key to practical widespread application of clinical
telemetry.

On-line, real-time computer monitoring of any medical signal is now possible. For
example, pacemaker performance can be monitored and checked by telephone. The
computer’s analysis can include English interpretations for any required time period. It
can include a comparison with any other required time period, and can also have
predictive statistical interpretations attached to it. Further means to trigger alarm displays
can be provided by real-time on-line computer monitoring.

Need for Clinical Design Foremost in designing a biomedical telemetry system should
be whether the data will be for clinical purposes of on-the-spot action, or for further
research. Research and clinical systems are vastly different and should always be
considered so. There have been far too many systems oriented toward the researcher and
sold to the clinician. There is a vast need for clinical and nursing telemetering aids,
specifically designed to meet clinical requirements.

If service systems include recording for storage, they can facilitate research as the result
of clinical experience. This mode of research has been stressed too little in the past and
may, in fact, be the most important of all. The laboratory of the future may be the clinical
ward rather than the isolated and too unrealistic laboratory.
 ON THE TIMING PROBLEM IN OPTICAL PPM COMMUNICATIONS*

R. M. GAGLIARDI
Department of Electrical Engineering
University of Southern California.

Summary The use of digital transmission with narrow light pulses appears attractive
for data communications, but carries with it a stringent requirement on system bit timing.
In this paper we investigate the effects of imperfect timing in a direct detection (non-
coherent) optical system using PPM bits. Particular emphasis is placed on specification
of timing accuracy, and an examination of system degradation when this accuracy is not
attained. Bit error probabilities are shown as a function of timing errors, from which
average error probabilities can be computed for specific synchronization methods. Of
significant importance is the presence of a residual, or irreducible error probability, due
entirely to the timing system, that cannot be overcome by the data channel.

Introduction The ability to generate extremely narrow, high energy light pulses from a
laser source has made the optical transmission of digital data extremely attractive for
modern communications. This possibility has fostered an exhaustive exploration of
optical communication systems, from both a theoretical and hardware point of view (e.g.
see [1]). The use of digital transmission with narrow pulses, however, carries with it an
extremely stringent requirement on system bit timing- i.e., time control of the system
sampling and integration intervals during each data bit. For the most part past analytical
studies have assumed perfect system timing, and the degradation caused by timing errors
in optical systems have been virtually ignored. In this paper, we investigate the effects of
imperfect timing in a direct-detection (non-coherent) optical communication system,
with particular emphasis on the specification of timing accuracy, and an examination of
the system degradation when this accuracy is not attained.

Consider an optical digital system that operates by transmitting a burst of energy in one
of two T sec adjacent time intervals to represent a binary bit. The above represents a two
level pulse position modulated (PPM) mode of transmission and is known to be optimal
under various criterion, when constrained in average transmitter power [2]. We shall
assume the transmitter and receiver operate diffraction-limited, so that the transmitted
energy corresponds to optical energy in a single spatial mode of the optical beam. Note
that T is the energy pulse width in time, 2T is the bit interval, and information is being

*
This work was sponsored by the National Aeronautics and Space Administration under NASA
Contract NGR-05-018-104.