Download H. Hutten -

Ventricular intramyocardial electrograms and their expected potential for
cardiac risk surveillance, telemonitoring and therapy management
H. Hutten
Graz University of Technology
Abstract – Ventricular intramyocardial electrograms are
recorded with electrodes directly from the heart either in
intraventricular or epimyocardial position and may be acquired
either from the spontaneously beating or from the paced heart.
The morphology of these signals differs significantly from that of
body surface ECG recordings. Although the morphology depends
on individual impacts like the electrode position, the volume and
anatomical shape of the heart and its changes during contraction,
the dimensions of the fibrotic capsule forming around the
electrode during the process of ingrowing etc, there is increasing
evidence that intramyocardial electrograms have a high potential
for cardiac risk surveillance, e.g. for arrhythmia detection,
recognition of rejection events in transplanted hearts, and
assessment of hemodynamic performance. Employing implants
with telemetric capabilities renders possible permanent and even
continuous cardiac telemonitoring. Furthermore the signals can
be utilized for supporting therapy management, e.g. in patients
with different kinds of cardiomyopathies. This paper shall
demonstrate some preliminary results and discuss the expected
potential. It will further discuss the problems of individual
evaluation which requires approaches like personalized
referencing based on similarity averaging and model-based signal
interpretation.
I.
System or the increasing incidence of ventricular ectopic beats.
Therefore the early recognition of such beginning episodes is
of fundamental relevance. A suitable procedure for cardiac
surveillance should allow quasi-continuous and in most cases
permanent, i.e. lifelong, application. The used equipment
should not constrain the patient in his daily routine activities
and preferably not require special handling by the patient, but
be forgettable if no risky situation is developing.
There is increasing preliminary evidence that intramyocardial electrograms have the potential for efficient cardiac risk
surveillance, especially if the system includes telemonitoring
features, e.g. based on the Bluetooth and WLAN technology,
cellular phone networks, intra- or internet technology or other
global communication technologies which are already
available, as well as advanced signal and information
processing combined with powerful database systems.
The most stringent constraint until now is the limited
processing capacity of the microprocessors which are used in
implants, usually 16-bit devices and in many cardiac
pacemakers still 8-bit devices. If more tasks have to be
accomplished, e.g. pacing, then the total load for the battery
may be another limiting constraint.
INTRODUCTION
II.
Diseases of the cardiovascular system are the major cause
for morbidity, short term inability to work, lifelong invalidity,
death and especially for health care expenditures in industrialized countries. Diagnostic terms like heart failure, cardiac
dysfunction, myocardial infarction, cardiomyopathy, cardiac
insufficiency, arrhythmia, and angina pectoris have become
well-known to many people, primarily people older than 65
years. These diseases seem to be not longer confined to
countries with a high living standard, but become an increasing
problem also in countries with poor economy. Typical patients
with a high risk potential for dramatic cardiac events are
patients on dialysis, with kidney insufficiency, diabetes
mellitus, high blood pressure, adiposity and certain lifestyle
factors.
Although many preventive programs have been proposed,
the results until now are not really convincing. There is some
evidence that programs aiming for changing the lifestyle do
not find wide acceptance.
Obviously individuals need the clear demonstration of their
actual situation combined with a timely warning that a sudden
event with high-risk potential may develop, e.g. the development of arrhythmia caused by a misbalance of the sympathetic
and parasympathetic activity of the Autonomous Nervous
METHODS
A. Signal Acquisition
Intramyocardial electrograms can be recorded with electrodes directly from the heart, either from the spontaneously
beating or from the paced heart. These signals are usually
utilized for the control of implanted cardiac pacemakers, e.g.
the inhibition of pacing and the control of timing intervals. The
intramyocardial electrograms have an amplitude of some
millivolts and a frequency range between dc and about 150 Hz
for the signals from the spontaneously beating heart and
between dc and about 50 Hz for paced events. They are
recorded usually in unipolar mode with the different electrode
either in epimyocardial or intraventricular position and the
pacemaker housing in some distance as the indifferent
electrode. If the same electrode is employed for both pacing
and recording of the Ventricular Evoked Response (VER), the
electrode must have low polarization effect and short time
constants for repolarization after pacing. These conditions are
fairly well met by electrodes with artificially enlarged surfaces,
e.g. by fractally coated or sputtered electrodes. Fig. 1
illustrates schematically the recording of intramyocardial
electrograms with a dual-channel device, e.g. a DDD-mode
pacemaker. Both channels may be used for pacing and/or
recording. If electrode E1 is used for pacing, electrode E2 can
be used to record the electrogram originating from electrode
E1. Those signals aquired with the non-pacing electrode are
frequently called VERX. The signal morphology of VERX
signals ranges between those of VERs and those of spontaneous events, depending on the distance between the two
electrodes E1 and E2. The analysis of VERX signals does not
only allow to calculate the mean propagation velocity between
electrode E1 (with proper consideration of the capture volume)
and E2 by the delay between the stimulation in E1 and the
appearance in E2, but may reveal histological information [1],
e.g. about the fiber architecture between the two electrodes.
However averaging is a proven method to improve the signalto-noise ratio and should preferably be used whenever
applicable. Usually sequences of 1 minute are recorded and
evaluated. Such sequences frequently contain not only VERs,
but also spontaneous events and fusion beats of different
degree. Fusion beats can show any signal morphology between
that of spontaneous events and that of paced events. For that
reason, a special procedure has been developed that uses
similarity averaging instead of time averaging for ensemble
averaging. It allows additionally to classify fusion beats by the
degree of their deviation from the pure VER morphology. This
renders possible some inference where the two excitation
wavefronts have met between the pacing electrode and the
origin of the spontaneous event. Fig. 3 shows a sequence of
intramyocardial electrograms that presents paced events,
spontaneous events and fusion beats as mixtures between these
two types of different degree.
Fig. 1: Schematic arrangement for recording intramyocardial electrograms
with a dual-channel device. SVE: Spontaneous Ventricular Event, VER:
Ventricular Evoked Response, VERX: Ventricular Evoked Response Crossed
B. Signal Processing
The signal morphology of intramyocardial electrograms as
near-distant potential recordings is significantliy different from
that of surface electrograms which are actually remote
projections of the volume signal source in a certain plane, e.g.
the usual bipolar EINTHOVEN leads in the frontal plane (fig. 2).
Fig. 3: Sequence of intramyocardial electrograms that contains paced
events (P), spontaneous events (S) and fusions of different degree (F).
Signal processing whether aiming for RR-interval times
series or for averaging requires the reliable and precise
identification of a certain fiducial time point for each event.
This is a rather simple task for paced events, however a
challenge for spontaneous events and especially for fusion
beats with their great variety in signal morphology. Hence, the
detection algorithm must fulfill challenging requirements
despite the limited performance of microprocessors in
implants. The zero-crossing algorithm seems to be a promising
candidate, but more advanced algorithms are required [2].
C. Signal Interpretation
Fig. 2: Typical signal morphology of intramyocardial electrograms
obtained from the spontaneously beating (a) and the paced heart (b) in
comparison with body surface electrograms.
Intramyocardial electrograms are less affected than surface
recordings by such disturbances as the relative displacement of
the heart against the electrode position caused by breathing, the
changes in impedance and distance between the heart and the
electrode due to different alveolar air filling, and muscle action
potentials. Therefore evaluation of single events is possible.
The signal morphology of intramyocardial electrograms
significantly depends on the exact positioning of the electrode
on or in the heart, on the volume and geometric shape of the
heart as well as on the volume changes during contraction, and
additionally about the growing of the fibrotic capsule around
the electrode after implantation. For that reason the signal
morphology of intramyocardial electrograms depends on
individual factors, although the fundamental shape is
comparable. If it is necessary to detect minor changes in the
signal morphology, this can be reached by two methods:
1. The patient is used as “his own reference”. If signals can
be acquired from the patient when he is in stable “normal”
condition, they can subsequently be used to recognize
deviations. Such deviations can be employed to identify their
meaning for further examinations by statistical comparison
with other symptoms.
2. A model-based approach is used to understand certain
details of the signal morphology. Such models should be as
realistic as possible, especially with regard to the geometry of
the individual heart. If the whole signal is considered, i.e. both
the depolarization and the repolarization phase, the model
should also be capable to simulate the contraction. For that
reason, 3D-images of the heart in enddiastolic and endsystolic
state should be available and the position of the electrode (or
of both electrodes) should be known as precisely as possible.
Appropriate models for excitation generation and spreading as
well as for contraction are now under development and may
become available for individual matching in the near future [2,
3, 4, 5, 6, 7, 8, 9, 10, 11, 12].
D. Signal Transmission
Remote cardiac risk surveillance requires the transmission
of the relevant information which may be the whole sequence
of intramyocardial electrograms or only an extracted risk
parameter either to an extracorporeal station in order to warn
the patient or to a central station which is in charge for the risk
management. The transmission of the rather low-bandwidth
electrograms (dc – 150 Hz) is not really a problem. The most
serious problem until now is the limited battery power in the
implant. For that reason transmission in available implants is
restricted to the transmission either of an alarm signal only or
to a “time window of interest” when the beginning of a risky
situation is detected. In the CHARM project signal transmission has been accomplished after activation by inductive
coupling which renders possible bi-directional transmission,
i.e. to interrogate the actual pacemaker status [13]. Inductive
coupling, however, will not be suitable for permanent
monitoring. Fig. 4 shows a schematic illustration of the
CHARM telemonitoring system.
extracorporeal receiver station has to be limited with regard to
the battery problem. But despite this constraint the mobility of
the patient should not be restricted. At present only distances
of about 2 m are acceptable. In future this problem may be
solved by body-worn devices which are integrated into the
clothing. These devices with both receiving and sending
features may transmit the data via Bluetooth or wireless LAN
technology to another near-by station that feeds the data into a
global network system like the internet, the cellular phone
network (e.g. Global System for Mobile Communications
GSM or Universal Mobile Telecommunication System UMTS)
or comparable technologies [14].
III.
APPLICATIONS AND RESULTS
A. Heart Transplant Rejection
Heart transplant patients frequently suffer by rejection
episodes. Preventive therapy is based on the application of
immunosuppressive drugs with the consequence of enhanced
risk for infections. Endomyocardial biopsy is the most widely
accepted surveillance method which, however, is an invasive
approach that can neither be standardized nor be applied for
daily routine examinations and is rather expensive. In the
CHARM-study (CHARM = Computerized Heart Acute
Rejection Monitoring) it has been proven that evaluation of
VERs allows reliable monitoring and shows good correlation
with the results of endocardial biopsy. The patients are
supplied with a pacemaker with broad-bandwidth telemetry for
signal transmission to an extracorporeal data acquisition
station. After acquisition the signals are transmitted via the
internet to a central processing station in Graz [13, 15]. Fig. 5
illustrates a typical VER from a patient during normal cardiac
condition, rejection and infection.
Fig. 5: VERs from a heart transplant patient with the heart in normal
condition, during a transient rejection episode and during an infection episode
(pod = post-operative day).
Fig. 4: Schematic illustration of the CHARM telemonitoring system [13]
In any case, the transmission range from the implant to the
B. Heart Rate Variability
Heart rate variability is well accepted for risk stratification
in patients. It is usually based on the determination of
consecutive RR-intervals with subsequent evaluation either in
the time or in the frequency mode. Details aiming for
standardization have been published in 1996 by the Task Force
of the European Society of Cardiology and the North
American Society for Pacing and Electrophysiology [16]. The
most urgent topic of this method is the assessment of the
sympathetic and parasympathetic control of cardiovascular
activity, primarily by the determination of the ratio of the
spectral power density in the low and in the high frequency
band. This allows also the assessment of the baroreflex
sensitivity [17]. A shortcoming of this approach is that the
evaluation requires more than 1 minute of steady state
conditions, i.e. it can not be applied to short episodes of HRV
fluctuations which seem to be relevant for continuous risk
assessment. Therefore a special procedure (Multimodal
Iterative Sinus Approximation) has been developed that
considers appropriately all clinical experience as requested by
the Task Force and allows to follow rapid HRV fluctuations
[18, 19]. A simplified version of this algorithm can be
implemented in a 16 bit-microprocessor in an implantable risk
monitor [20].
Fig. 6 depicts the results obtained with this approach in the
bimodal version for the evaluation of the fetal heart rate with
this approach in the bimodal version. The rapid fluctuations of
the spectral power density in the low frequency (LF) and in the
high frequency (HF) band as well as the fluctuations of the 2
frequencies with the highest power density in these two
frequency ranges demonstrate the superiority of this approach
over the usual procedures based on Fast Fourier Transformation and Autoregressive Models. In this case the unborn
baby was healthy, the fetal heart rate showed regular
accelerations and decelerations [20].
baroreflex activity and thereby the Autonomous Nervous
System with its feedback on the cardiac activity. This effect
can be assessed globally with the parameters of the heart rate
turbulence describing the dynamic transition of heart rate after
an ectopic beat for about 20 s [21]. Another risk caused by
ventricular ectopic beats may occur if they fall into the
vulnerable period of the preceding beat.
Fig. 7 illustrates the frequency of ventricular ectopic beats
in a patient during hemodialysis [22]. The high rate of such
ectopic beat has significant cardiovascular consequences.
However, these patients are under high risk also at home,
especially in the time after leaving the hospital and may benefit
from continuous risk monitoring.
Fig. 7: Time course of ventricular ectopic beats in a patient during
hemodialysis. Each time segment corresponds to 5 minutes
D. Hemodynamic Performance
Intramyocardial electrograms are electrophysiological
signals that represent the spatio-temporal spreading of the
depolarizing and repolarizing wavefronts in the myocardium.
Fig. 6: a. Records of fetal heart rate (upper part) and tocogram (lower
part), b. HRV parameters LF Band, HF Band and characteristic frequencies
obtained by application of the Bimodal Iterative Sinus Approximation [20]
C. Ventricular Ectopic Beats
Ventricular ectopic beats have some remarkable consequences. (1) The ventricular filling volume is reduced due to
the shortened filling phase. (2) The atrial contraction is not
preceding the ventricular contraction in proper time and hence
does not effectively contribute to ventricular filling. (3) The
spatio-temporal excitation spreading and, consequently the
ventricular contraction sequence differs from that of a regular
excitation originating from the Sinus Node and propagating
along the specific conduction system. All these impacts result
in a reduced left-ventricular stroke volume with a sudden
decrease of the arterial blood pressure that stimulates the
Fig. 8: Evaluation of the VER-parameter DAmp for 6 heart transplant
patients in lying and supine position.
The morphology of these signals, however, depends also on
the geometric shape and volume of the heart, i.e. the
enddiastolic and the endsystolic volume, and consequently
monitors the stroke volume.. This dependence can in principle
demonstrated using appropriate models. Until now, however,
only very coarse models without sufficient individual matching
have been made available. But in recent past that dependence
could additionally be confirmed in patients by comparison
with thermodilution and echocardiographic measurements and
also by measurements during orthostatic tests and workload
challenges [23, 24, 25]. It could even be demonstrated that
parameters extracted from intramyocardial electrograms
correlate with the NYHA classification and, hence, allow to
monitor the efficiency of therapy management. Fig. 8 shows
the results obtained from patients who underwent orthostatic
tests.
E. Therapy Management
There is preliminary and challenging evidence that intramyocardial electrograms may be useful for supporting
therapeutic procedures and therapy management. The
fundamentals are based on the analysis and understanding of
the signal morphology of fusion beats [26, 27]. Fusion beats
can have different causes, but finally result if two depolarizing
wavefronts meet with the consequence of mutual annihilation,
i.e. none of the wavefronts proceeds beyond the meeting
border line. These two wavefronts can originate from different
sources, e.g. a spontaneous excitation from the Sinus Node
meets with a paced event or with an ectopic beat, or the same
excitation propagates on different pathways to the same
syncytium, e.g. the excitation arrives in the ventricles via the
HIS bundle and an accessory pathway. In pacemaker therapy
the avoidance of fusion beats is of high relevance, both in
order not to waste battery charge and also not to disturb the
timing protocol.
In the recent past the Ventricular Resynchronization
Therapy has gained increased attention. It is based on
multisite-pacing, e.g. biventricular pacing. The provoked
fusion beats are followed by a changed spatio-temporal
contraction as compared with regular excitation spreading.
Interpretation of the intramyocardial electrograms representing
fusion beats caused by multisite pacing may help to optimize
the hemodynamic performance of the heart, e.g. by matching
the delay between the two pacing pulses.
In patients with Hypertrophic Obstructive Cardiomyopathy
(HOCM) the contraction starts too early at the aortic outflow
and thereby enlarges the outflow resistance before the ejection
of blood by ventricular contraction begins. This enlarged
outflow resistance causes hypertrophy of the ventricular
muscle. The therapeutic concept utilizes a dual-chamber
pacemaker that initiates ventricular contraction with sufficiently short atrio-ventricular interval (AVI) after sensing an atrial
excitation in order to avoid the premature contraction at the
aortic outflow.
Fig. 9 illustrates the change in the morphology of the
intramyocardial electrogram for different AVIs and the
resulting impact on the Left Ventricular Outflow Tract
Gradient (LVOTG) which is a measure for the aortic outflow
resistance [28].
Fig. 9: Impact of different atrio-ventricular intervals (AVI) on the signal
morphology of intramyocardial electrograms and the Left Ventricular Outflow
Tract Gradient (LVOTG) in a patient
Fig. 10 depicts intramyocardial electrograms acquired from
two patients with dilative cardiomyopathy (i.e. candidates for
heart transplantation) and from one patient after heart
transplantation which show distinct differences in amplitude
and morphology. These deviations may be caused primarily by
differences in the stroke volume (or ejection fraction which
usually in patients with dilative cardiomyopathy is much lower
than in individuals with normal cardiac situation), but it may
also be due to the enlargement of the heart silhouette and even
the reduction in the thickness of the ventricular wall. Model
based simulations have shown that these two variables can also
affect the signal morphology.
Fig. 10: Intramyocardial electrograms recorded from two patients with
dilative cardiomyopathy and one patient after heart transplantation.
F. Other applications
Fig. 11 illustrates intramyocardial electrograms acquired
from a heart transplant patient paced with different pacing rate.
With higher pacing rate the deviation in the amplitude of the
intramyocardial electrogram, especially in the period when
contraction starts and the signal represents the enddiastolic
volume, is significant and regular. This may be explained by a
reduction in the enddiastolic filling volume. This finding may
cause some speculations whether the effect can be utilized to
develop a completely new generation of rate-adaptive cardiac
pacemakers. Independent whether the adjustment of the stroke
volume to a change in the hemodynamic requirement is
transmitted primarily as inotropic information via the
Autonomous Nervous System or by the FRANK-STARLING
mechanism as in denervated hearts like in heart transplant
patients, if it can be extracted, then it can be utilized for rate
adaptation.
Fig. 11: Recordings of intramyocardial electrograms acquired from a heart
transplant patient during pacing with different pacing rates.
IV.
DISCUSSION
Permanent and continuous monitoring of high-risk patients
with cardiac failures as well as computer-assisted therapy
management is an urgent problem in most industrialized
countries and becoming an urgent problem also in other
countries with poorer economy. Advanced information and
communication technology is already available that renders
possible the establishment of global networks for patient
surveillance. The most difficult problem seems to be the
identification and acquisition of the most suitable information,
especially regarding the earliest detection of beginning lifethreatening events. Intramyocardial electrograms may open a
new access to cardiac monitoring with regard to arrhythmia
detection, recognition of dangerous ventricular ectopic beats,
episodes of rejection in transplanted hearts, and other diseases
resulting in cardiac insufficiency, but also with regard to the
support of therapy management by drug application and
adjustment of procedures like multisite pacing. A problem,
however, that is not satisfactorily solved until now and
requires more investigation is the understanding and
interpretation of the signal morphology in order to extract
reliable fiducial parameters and to “personalize” monitoring
and therapy management. A promising approach is modelbased interpretation utilizing personalized 3D data sets [29].
ACKNOWLEDGMENT
This work has in part been supported by the Austrian
Science Fund in the Project No. P16965-N04 what is gratefully
acknowledged.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Kraft, A., Schreier, G., Hutten, H., et al.: Assessment of Ventricular
Excitation Propagation Velocity using Intramyocardial Electrograms
and Electron Beam Tomography. Medical & Biological Engineering &
Computing 37, Suppl. 2, Part II: 994 – 995 (1999)
Stiegmaier, W., Hutten, H., Rauchegger, G.: Implementation of an
Efficient R-Peak Detection Algorithm on a 16 bit Microprocessor for a
Subcutaneously Implantable Heart Risk-Monitor. IFMBE Proc. 11(1):
Proc. 3rd European Medical and Biological Engineering EMBEC’05
(ISSN 1727-1983): Paper No. 1611 (2005)
LeGrice, I.J., Hunter, P.J., Small, B.H.: Laminar structure of the heart: a
mathematical model. J. Physiol. 272: H2466 – H2476 (1997).
LeGrice, I.J., Hunter, P.J., Young, A.A., Samll, B.H.: The architecture
of the heart: a data-based model. Phil. Trans. R. Soc. A 359 : 1217 –
1232 (2001).
Smith, N.P., Mulquiney P.J., Nash, M.P et al . : Mathematical modelling
of the heart: cell to organs. Chaos, Solitons, Fractals 13: 1613 – 1621
(2001).
Kohl, P., Noble, D., Hunter, P.J. (eds.): The integrated heart: modelling
cardiac structure and function. Phil. Trans. R. Soc. A 359 (2001).
Noble, D.: Modelling the heart: from genes to cells to the whole organ.
Science 295: 1678 – 1682 (2002).
Winslow, R.L., Helm, P., Baumgartner jr., W. et al.: Imaging-based
integrative models of the heart: closing the loop between experiment
and simulation. In: Bock, G., Goode, J.A. (eds.): “In silico” – Simulation of biological processes. Hoboken, John Wiley & Sons, Inc. : 129 –
140 (2002).
Noble, D.: The heart cell in silic: success, failures and prospects. In:
Bock, G., Goode, J.A. (eds.): “In silico” – Simulation of biological
processes. Hoboken, John Wiley & Sons, Inc. : 182 – 1^94 (2002).
[10] Sachse, F.B.: Computational Cardiology: Modeling of Anatomy,
Electrophysiology, and Mechanics. LNCS 2966. Heidelberg, Springer
Press (2004).
[11] Seemann. G.: Modeling of electrophysiology and tension development
in the human heart. PhD thesis, Institut fuer Biomedizinische Technik,
Universitaetsverlag Karlsruhe ISBN 3-937300-66-X (2005).
[12] L. Fritz, A. Prassl, H. Hutten et al.: Model-Based Analysis and
Interpretation of Human Epicardial Electrograms using a Bidomain Slab
Model. Proc. 3rd European Medical & Biological Engineering Conference EMBEC’05 2005:Paper No 2125 (2005)
[13] H. Hutten, G. Schreier, P. Kastner et al.: CHARM – Computerized
Heart Acute Rejection Monitoring. Biomedizinische Technik 41: 35 – 40
(1996).
[14] Hutten, H., Richnovsky, D., Staber, G., Rauchegger, G.: Wireless ECG
home monitoring system based on Bluetooth and GSM / GPRS. Electronics in Medicine 2004 (Cardiostim): St. Petersburg (Russia) 2004.
[15] B. Grasser, B. Iberer, G, Schreier et al.: Computerized Heart Allograft
Recipient Monitoring (CHARM) – a multicenter study. J. Heart Lung
Transplant. 20: 214 (2001).
[16] Task Force of the European Society of Cardiology and The North
American Society of Pacing and Electrophysiology: Heart rate variability – Standards of measurement, physiological interpretation and clinical use. Eur. Heart J. 17: 354 – 381 (1996).
[17] Fahn, B., Scharfetter, H., Wirnsberger, G., Hutten, H.: Spectral Analysis
of Heart Rate Variability during Hemodialysis. Medical & Biological
Engineering & Computing 37, Suppl. 2, Part II: 1280 – 1281 (1999)
[18] Rauchegger, G., Hutten, H.: Kurzzeit-Herzratenvariabilitätsanalyse
durch multimodale iterative Sinusapproximation. Biomedizinische
Technik 48 (Ergänzungsband 1): 166 – 167 (2003).
[19] Rauchegger, G., Hutten, H: Short-Time Heart-Rate-Variability Analysis
with Recursive Multimodal Iterative Sine-Approximation. IFMBEProceedings, Vol. 6, 2004: MEDICON and HEALTH TELEMATICS
2004; CD-Ausgabe, ISSN: 1727-1983, ISBN: 88-7780-308-8 (2004).
[20] Rauchegger, G.: Multimodale Iterative Sinusapproximation (MIS) –
Eine dynamische Zeit-Frequenz-Transformation zur kardialen
Risikostratifizierung. PhD-Thesis Graz University of Technology,
(2004).
[21] Schmidt, G., Malik, M., Barthel, P. et al.: Heart-rate turbulence after
ventricular premature beats as a predictor of mortality after acute myocardial infarction. The Lancet 353: 1390 – 1396 (1999).
[22] Andorfer, C., Hutten, H.: Kardiovaskuläres Risikomonitoring während
der Hämodialyse durch EKG-Auswertung. Biomedizinische Technik 48
(Ergänzungsband 1): 494 – 495 (2003).
[23] H. Hutten, P. Kastner, G. Schreier: Hemodynamic assessment by
evaluation of intramyocardial electrograms. Proc. 20th Ann. Int. Conf.
IEEE-EMBS Hongkong: 395 – 298 (1998).
[24] Hribernigg, M., Ebner, E., Krätschmer, H., Hutten, H., Rauchegger, G.:
Correlation Between Electrophysiologic and Hemodynamic Parameters
of the Paced Heart. Proc. 2nd European Medical and Biological Engineering Conference EMBEC’02, Vol. 3 / 2002 in the IFMBE Proceedings Series, Part. I: 336 – 337 (2002).
[25] Hribernigg, M., Ebner, E., Krätschmer, H., Hutten, H., Rauchegger, G.:
Utilization of Intramyocardial Electrograms for Monitoring the Cardiac
Performance in Accordance with the NYHA Classification. Proc. 2nd
European Medical and Biological Engineering Conference EMBEC’02,
Vol. 3 / 2002 in the IFMBE Proceedings Series, Part. I: 374 – 375
(2002).
[26] H. Hutten, B. Tschapeller, L. Fritz: Some interesting Properties of
Cardiac Fusion Beats. Proc. World Congress on Medical Physics and
Biomedical Engineering 2006: 3303 – 3306 (2006).
[27] L. Fritz, H. Hutten: Analysis of the Impact of Fusion Beats onto
Epicardial Electrograms Based on a Bidomain Slab Model. Proc. World
Congress on Medical Physics and Biomedical Engineering 2006: 3295
– 3298 (2006).
[28] Schreier, G., Kastner, P., Hutten, H. et al.: Intramyocardial Electrogram
Analysis for Hemodynamic Optimization in Patients Paced for Hypertrophic Obstructive Cardiomyopathy. Medical & Biological Engineering & Computing 37, Suppl. 2, Part I: 704 – 705 (1999)
[29] Hutten, H., Tschapeller, B., Fritz, L., Rauchegger, G.: Model-Based
Interpretation of Ventricular Evoked Responses. Proc. 17th Biennial
Int. EURASIP Conf. BIOSIGNAL 2004. Jan J. Kozumplik, J., Provaznik,
I. (eds.), Vutium Press (ISSN: 1211-412X), Brno University of Technology: 44 – 46 (2004).