US8929984

USOO8929.
984B2
(12) United States Patent (10) Patent No.: US 8,929,984 B2

Ghosh et al. (45) Date of Patent: Jan. 6, 2015
(54) CRITERIA FOR OPTIMAL ELECTRICAL 6,968,237 B2 * 1 1/2005 Doan et al. .................... 607,122
RESYNCHRONIZATION DURING FUSION 7,031,777 B2 : 4/2006 Hine et al. ..... 607,122
PACING 7,181,284 B2 2, 2007 Burnes et al. ................... 6O7/25
7,313,444 B2 * 12/2007 Pianca et al. ... ... 607,126
7,860,580 B2 * 12/2010 Falk et al. ... ... 607 127
(71) Applicant: Medtronic, Inc., Minneapolis, MN (US) 7.917,214 B1* 3/2011 Gill et al. ....... ... 607.9
8,036,743 B2 * 10/2011 Savage et al. .......... 6O7.5
(72) Inventors: Subham Ghosh, Blaine, MN (US); 8,145,308 B2 * 3/2012 Sambelashvili et al. 607/25
Robert W Stadler, Shoreview, MN (US) 8,160,700 B1 * 4/2012 Ryu et al. ............... ... 607.9
8,615,298 B2 * 12/2013 Ghosh et al. ... 607/28
8,620,433 B2 * 12/2013 Ghosh et al. ....... 607/28
(73) Assignee: Medtronic, Inc., Minneapolis, MN (US) 2006/0235478 A1* 10, 2006 Van Gelder et al. .............. 6079
- 2009.0036947 A1 2/2009 Westlund et al. ............... 607/42
(*) Notice: Subject to any disclaimer, the term of this 2009,0270729 A1* 10, 2009 Corbucci et al. ... ... 600,438
patent is extended or adjusted under 35 38.933 A. 1938: E. et etal.al...
Osenberg ............. 2. 2
U.S.C. 154(b) by 0 days. 2011/0098772 A1* 4/2011 Min ................................ 607/28
2011/0137369 A1* 6/2011 Ryu et al. ........................ 6O7/27
(21) Appl. No.: 13/772,890 2013/0190834 A1* 7, 2013 Ghosh et al. .................... 6O7/17
2013/0218223 A1* 8, 2013 Ghosh et al. .................... 607/28
(22) Filed: Feb. 21, 2013 (Continued)
(65) Prior Publication Data Primary Examiner — Niketa Patel
US 2014/O236253 A1 Aug. 21, 2014 Assistant Examiner — Christopher A Flory
(74) Attorney, Agent, or Firm — Carol F. Barry
(51) Int. Cl.
A61N L/37 (2006.01) (57) ABSTRACT
A6 IN L/368 (2006.01) Generally, the disclosure is directed one or more methods or
(52) U.S. Cl. systems of cardiac pacing employing a plurality of left ven
CPC ............ A61N I/3706 (2013.01); A61N I/3686 tricular electrodes. Pacing using a first one of the left ven
(2013.01) tricular electrodes and measuring activation times at other
USPC .................................... 607/28: 607/9: 607/25 ones of the left and right Ventricular electrodes. Pacing using
(58) Field of Classification Search a second one of the Ventricular electrodes and measuring
CPC. A61N 1/3622; A61N 1/3682: A61N 1/3686; activation times at other ones of the left ventricular elec
A61N1A37O6 trodes. Employing weighted Sums of the measured activation
USPC .................................................. 607/9, 25, 28 times to measure a fusion index and select one of the left
See application file for complete search history. Ventricular electrodes for delivery of Subsequent pacing
pulses based on comparing fusion indices during pacing from
(56) References Cited different LV electrodes. One or more embodiments use the
U.S. PATENT DOCUMENTS same fusion index to select an optimal A-V delay by compar
ing fusion indices during pacing with different A-V delays at
4,497.326 A * 2/1985 Curry ............................ 607,123 resting atrial rates as well as rates above the resting rate.
5,443,492 A * 8/1995 Stokes et al. .................. 607 131
5,628,778 A * 5/1997 Kruse et al. ................... 607,123 36 Claims, 10 Drawing Sheets
-i.
(e)
Paceata SAWEPAW firstly
electrode U
104
Measure activation times atother
ones of Welectrodes
106
Determine Weighted fusion index
for first electrode
Pace at same SAWFPAW

second Li electrode
Measure activation times atother

ones of Welectrodes U
12
Determine weighted fusion index
for secondly electrode Ll/
Select optimal
LW electrode L
US 8,929,984 B2
Page 2
(56) References Cited 2013/0218225 A1 8, 2013 Ghosh et al. .................... 607/28

2013/0218226 A1* 8, 2013 Ghosh et al. .................... 607/28
U.S. PATENT DOCUMENTS 2013/0218227 A1* 8, 2013 Ghosh et al. .................... 607/28
2013/0218224 A1* 8, 2013 Ghosh et al. .................... 607 28 * cited by examiner
U.S. Patent Jan. 6, 2015 Sheet 1 of 10 US 8,929,984 B2
PROGRAMMER
24
-"
14
S
cy
40 42 44 45 46 47 62 64 48 50 66 58
H
SWITCH MODULE
85.
THERAPY DELIVERY
SENSING MODULE MODULE
86. 84
TELEMETRY
MODULE PROCESSOR MEMORY
88 80 82.
POWER
SOURCE
90
-
102
Pace at a SAVIPAV first LV
electrode
104
MeaSure activation timeSat Other
OneS Of LV electrodes
106
for first LV electrode
108
Pace at Same SAVIPAV
SeCOnd LV electrode
110
MeaSure activation timeSat Other
OneS Of LV electrodes
112
for SeCOnd LV electrode
114
Select optimal
LV electrode
Fi 9 . . 5
Measure activation times at 202 200

LV electrodes during RV
intrinsic rhythm
204
Store AT data
1 2O6
N = number of LW electrodes
T = Threshold
Determine Ffor One LV 208

electrode
210 Optimal electrode =
Determine F for another ?ering
LV electrode
234
Difference
in F valueSeSS
than T2
230
Other criteria
eliminates electrode
Eliminate One electrode
Fig. 6A
Measure activation times at 202 200

LV electrodes during RV Only
pacing Or intrinsic rhythm
204
Store AT data
1 2O6
N = number of LW electrodes
T = Threshold
Determine F for One LV 208

electrode
210 Optimal electrode =
Determine F for another remaining LV
electrode
LV electrode
234
Determine difference between
One and another FI data
Difference in F Values is
Compared to a threshold T
Difference Eliminate electrode

in F Values leSS with larger F
than T2
230
Compare One AT to another
AT
One S.-2,
Another
leSS
than another
Eliminate another
electrode
AT? AT
228
226
Yes 7-2
Adjust pacing parameters Eliminate One electrode
Fig. 6B
--
302
Pace LV
electrode at first A-V delay
304
MeaSure activation timeSat
LV electrodes
306
dyssynchrony for first A-V delay
308
PaCe LV electrode at SeCOnd
A-V delay
310
MeaSure activation timeSat
LV electrodes
312
for second A-V delay
314
Select optimal
A-V delay
Fi 9 7
BL
RVAT(1A) 400 mS WindoW

= tRVA -1 starting on atrial event
- 16 mV
0 mS 400 mS
+16 mV
LVAT(4A)
= t V4-A
0 mV LV4-RV COil
BL
atrial event Ventricular pace (at LV1)
Fig. 8
US 8,929,984 B2
1. 2
CRITERA FOR OPTIMAL ELECTRICAL utilized and structural changes may be made without depart
RESYNCHRONIZATION DURING FUSION ing from (e.g., still falling within) the scope of the disclosure
PACING presented hereby.
As described herein, a physician implanting a medical
FIELD device can use criteria, stored in a programmer, to automati
cally select optimized location(s) and/or parameters for deliv
The present disclosure relates to implantable medical ery of cardiac resynchronization therapy (CRT) through
devices (IMDS), and, more particularly, to selecting an opti fusion pacing. For example, in one or more embodiments,
mal left ventricular electrode on a medical electrical lead criteria can be used to determine an optimal left ventricular
extending from an IMD to deliver cardiac therapy. 10 (LV) electrode from which electrical stimuli is delivered to
the left ventricle. After the optimal LV electrode has been
BACKGROUND selected, other criteria can be used to optimize an atrioven
tricular delay for maximal cardiac resynchronization. In one
Implantable medical devices (IMD) are capable of utiliz or more other embodiments, different criteria can be used to
ing pacing therapies, such as cardiac resynchronization 15 determine an optimal right ventricular (RV) electrode from
therapy (CRT), to maintain hemodynamic benefits to which electrical stimuli is delivered to the right ventricle.
patients. Fusion pacing is a form of CRT therapy. Fusion After the optimal RV electrode has been selected, criteria can
pacing reduces the power consumed by an implantable medi be used to optimize an atrioventricular delay for maximal
cal device since only one ventricle is paced in coordination cardiac resynchronization. Implementation of teachings of
with the other ventricle's intrinsic activation. For example, this disclosure can potentially improve CRT response in
the left ventricle (LV) can be paced in coordination with the patients through fusion pacing. For example, heart failure
intrinsic right ventricle (RV) activation or vice versa. Recent patients with stable intrinsic A-V conduction and intraven
developments infusion pacing have been described in printed tricular conduction disorder (e.g. left bundle branch block,
publications. For example, US Patent Publication 2011/ right bundle branch block) may have an improved response to
0137639 by Ryu et al. discloses that the optimal left ventricle 25 CRT by implementing features described herein.
electrode is selected based upon conduction Velocities. Exemplary methods, devices, and systems are described
Another U.S. Pat. No. 7,917,214 to Gill et al. discloses that with reference to FIGS. 1-8. It is appreciated that elements or
the optimal left ventricle electrode is selected based upon processes from one embodiment may be used in combination
activation times and activation recovery interval dispersions. with elements or processes of the other embodiments. The
It is desirable to develop additional methods and systems to 30 possible embodiments of such methods, devices, and systems
optimize fusion pacing. using combinations of features set forth herein is not limited
to the specific embodiments shown in the Figures and/or
BRIEF DESCRIPTION OF THE DRAWINGS described herein. Further, it will be recognized that the
embodiments described herein may include many elements
FIG. 1 is a schematic diagram of an exemplary system 35 that are not necessarily shown to scale. Still further, it will be
including an exemplary implantable medical device (IMD). recognized that timing of the processes and the size and shape
FIG. 2 is a schematic diagram of the exemplary IMD of of various elements herein may be modified but still fall
FIG 1. within the scope of the present disclosure, although certain
FIGS. 3-3A are schematic diagrams of an enlarged view of timings, one or more shapes and/or sizes, or types of ele
a distal end of a medical electrical lead disposed in the left 40 ments, may be advantageous over others.
ventricle. FIG. 1 is a conceptual diagram illustrating an exemplary
FIG. 4 is a block diagram of an exemplary IMD, e.g., the therapy system 10 that may be used to deliver fusion pacing
IMD of FIGS 1-2. therapy to a patient 14 that may, but not necessarily, be a
FIG. 5 is a general flow chart of an exemplary method that human. Fusion pacing typically involves left ventricle (LV)
involves determining a weighted electrical dyssynchrony for 45 only pacing with an electrode on the LV medical electrical
selecting an optimal left ventricle electrode to pace a left lead in coordination with the intrinsic right ventricle (RV)
ventricle. activation. Alternatively, fusion pacing can involve pacing the
FIG. 6A is a general flow chart of an exemplary method RV with an electrode on the RV medical electrical lead in
that involves selecting an optimal electrode to pace a ven coordination with the intrinsic LV activation.
tricle. 50 The therapy system 10 may include an implantable medi
FIG. 6B is a general flow chart of another exemplary cal device 16 (IMD), which may be coupled to leads 18, 20.
method that involves selecting an optimal electrode to pace a 22 and a programmer 24. For the sake of brevity, programmer
ventricle. 24 includes a computer capable of the functions represented
FIG. 7 is a general flow chart of an exemplary method that in FIG. 4 that are incorporated herein.
involves determining a weighted electrical dyssynchrony for 55 The IMD 16 may be, e.g., an implantable pacemaker, car
selecting an optimal A-V delay. dioverter, and/or defibrillator, that provides electrical signals
FIG. 8 depicts a ventricular electrogram that includes ven to the heart 12 of the patient 14 via electrodes coupled to one
tricular activations times. or more of the leads 18, 20, 22.
The leads 18, 20, 22 extend into the heart 12 of the patient
DETAILED DESCRIPTION OF EXEMPLARY 60 14 to sense electrical activity of the heart 12 and/or to deliver
EMBODIMENTS electrical stimulation to the heart 12. In the example shown in
FIG. 1, the right ventricular (RV) lead 18 extends through one
In the following detailed description of illustrative embodi or more veins (not shown), the Superior Vena cava (not
ments, reference is made to the accompanying figures of the shown), and the right atrium 26, and into the right ventricle
drawing which form a parthereof, and in which are shown, by 65 28. The left ventricular coronary sinus lead 20 extends
way of illustration, specific embodiments which may be prac through one or more veins, the Vena cava, the right atrium 26,
ticed. It is to be understood that other embodiments may be and into the coronary sinus 30 to a region adjacent to the free
US 8,929,984 B2
3 4
wall of the left ventricle (LV) 32 of the heart 12. The right herein are generally disclosed in U.S. Pat. No. 7,031,777,
atrial (RA) lead 22 extends through one or more veins and the U.S. Pat. No. 6,968,237, and US Publication No. 2009/
vena cava, and into the right atrium 26 of the heart 12. 0270729, all of which are incorporated herein by reference in
The IMD 16 may sense, among other things, electrical their entirety. Moreover, U.S. Pat. No. 7,313,444, incorpo
signals attendant to the depolarization and repolarization of 5 rated by reference, discloses a LV pacing lead such that the
the heart 12 via electrodes coupled to at least one of the leads LV electrodes are about equally spaced, which could also be
18, 20, 22. In some examples, the IMD 16 provides pacing used to implement the present disclosure.
therapy (e.g., pacing pulses) to the heart 12 based on the In the illustrated example, bipolar or unipolar electrodes
electrical signals sensed within the heart 12. The IMD 16 may 40, 42 (also referred to as RV electrodes) are located proxi
be operable to adjust one or more parameters associated with 10 mate to a distal end of the lead 18. Referring briefly to FIGS.
the pacing therapy such as, e.g., pulse width, amplitude, Volt 3-3A, the electrodes 44, 45, 46 are located proximate to a
age, burst length, etc. Further, the IMD 16 may be operable to distal end of the lead 20 and the bipolar or unipolar electrodes
use various electrode configurations to deliver pacing 56, 50 (FIG. 2) are located proximate to a distal end of the lead
therapy, which may be unipolar or bipolar. The IMD 16 may 22. Electrodes 44, 45, 46 and 47 can be bipolar electrodes,
also provide defibrillation therapy and/or cardioversion 15 unipolar electrodes or a combination of bipolar and unipolar
therapy via electrodes located on at least one of the leads 18, electrodes. Additionally, electrodes 44, 45,46 and 47 have an
20, 22. Further, the IMD 16 may detect arrhythmia of the electrode surface area of about 5.3 mm to about 5.8 mm.
heart 12, such as fibrillation of the ventricles 28, 32, and Electrodes 44, 45, 46, and 47 are also referred to as LV1
deliver defibrillation therapy to the heart 12 in the form of (electrode 1), LV2 (electrode 2), LV3 (electrode 3), and LV4
electrical pulses. In some examples, IMD 16 may be pro (electrode 4), respectively. As shown, lead 20 includes a
grammed to deliver a progression of therapies, e.g., pulses proximal end 92 and a distal end 94. The distal end 94 is
with increasing energy levels, until a fibrillation of heart 12 is placed in or near LV tissue. Skilled artisans appreciate that LV
stopped. electrodes (i.e. left ventricle electrode 1 (LV1) 44, left ven
In some examples, a programmer 24, which may be a tricle electrode 2 (LV2)45, left ventricle electrode 3 (LV3)46,
handheld computing device or a computer workstation, may 25 and left ventricle 4 (LV4) 47 etc.) on lead 20 can be spaced
be used by a user, such as a physician, technician, another apart at variable distances. For example, electrode 44 is a
clinician, and/or patient, to communicate with the IMD 16 distance 96a (e.g. about 21 mm) away from electrode 45,
(e.g., to program the IMD 16). For example, the user may electrodes 45 and 46 are spaced a distance 96b (e.g. about 1.3
interact with the programmer 24 to retrieve information con mm to about 1.5 mm) away from each other, and electrodes 46
cerning one or more detected or indicated faults associated 30 and 47 are spaced a distance 96c (e.g. 20 mm to about 21 mm)
within the IMD 16 and/or the pacing therapy delivered there away from each other.
with. The IMD 16 and the programmer 24 may communicate The electrodes 40, 44, 45,46, 47, 48 may take the form of
via wireless communication using any techniques known in ring electrodes, and the electrodes 42, 47, 50 may take the
the art. Examples of communication techniques may include, form of extendable helix tip electrodes mounted retractably
e.g., low frequency or radiofrequency (RF) telemetry, but 35 within the insulative electrode heads 52, 54,56, respectively.
other techniques are also contemplated. Each of the electrodes 40, 42, 44, 45, 46, 47, 48, 50 may be
FIG. 2 is a conceptual diagram illustrating the IMD 16 and electrically coupled to a respective one of the conductors
the leads 18, 20, 22 of therapy system 10 of FIG. 1 in more (e.g., coiled and/or straight) within the lead body of its asso
detail. The leads 18, 20, 22 may be electrically coupled to a ciated lead 18, 20, 22, and thereby coupled to respective ones
therapy delivery module (e.g., for delivery of pacing therapy), 40 of the electrical contacts on the proximal end of the leads 18,
a sensing module (e.g., one or more electrodes to sense or 20, 22. The electrodes 40, 42, 44, 45, 46, 47, 48, 50 may
monitor electrical activity of the heart 12 for use in determin further be used to sense electrical signals attendant to the
ing effectiveness of pacing therapy), and/or any other mod depolarization and repolarization of the heart 12. The electri
ules of the IMD 16 via a connector block 34. In some cal signals are conducted to the IMD 16 via the respective
examples, the proximal ends of the leads 18, 20, 22 may 45 leads 18, 20, 22. In some examples, the IMD 16 may also
include electrical contacts that electrically couple to respec deliver pacing pulses via the electrodes 40, 42, 44, 45,46, 47.
tive electrical contacts within the connector block 34 of the 48,50 to cause depolarization of cardiac tissue of the patients
IMD 16. In addition, in some examples, the leads 18, 20, 22 heart 12. In some examples, as illustrated in FIG. 2, the IMD
may be mechanically coupled to the connector block 34 with 16 includes one or more housing electrodes, such as housing
the aid of set screws, connection pins, or another Suitable 50 electrode 58, which may be formed integrally with an outer
mechanical coupling mechanism. Surface of a housing 60 (e.g., hermetically-sealed housing) of
Each of the leads 18, 20, 22 includes an elongated insula the IMD 16 or otherwise coupled to the housing 60. Any of the
tive lead body, which may carry a number of conductors (e.g., electrodes 40, 42, 44, 45, 46, 47, 48 and 50 may be used for
concentric coiled conductors, straight conductors, etc.) sepa unipolar sensing or pacing in combination with housing elec
rated from one another by insulation (e.g., tubular insulative 55 trode 58. Further, any of electrodes 40, 42, 44, 45,46, 47, 48,
sheaths). Exemplary leads that can be useful for the present 50, 58, which are not being used to deliver pacing therapy,
disclosure include U.S. Pat. No. 5,922,014, U.S. Pat. No. may be used to sense electrical activity during pacing therapy
5,628,778, U.S. Pat. No. 4,497,326, U.S. Pat. No. 5,443,492, (e.g., for use in determining electrical activation times). Elec
U.S. Pat. No. 7,860,580 or US Patent Application trical activation time can be used to determine whether fusion
20090036947 filed Apr. 30, 2008 such that electrodes are 60 pacing produces effective contraction of the heart based on
added and/or spaced apart in a manner similar to that dis metrics of electrical dyssynchrony derived from the ventricu
closed in the figures of the present application, all of listed lar activation times.
patents and applications are incorporated by reference in their Electrical activation time or local electrical activity is
entirety. Additional lead and electrode configurations that determined relative to timing of a fiducial, an indicator of a
may be adapted for use with the present disclosure by adjust 65 global cardiac event (e.g. timing of contraction of a chamber
ing lead shape, length, electrode number and/or electrode to of the heart, timing of pacing of a chamber of the heart, etc.)
effectively avoid phrenic nerve stimulation as described For example, the fiducial may be the onset of the QRS waves
US 8,929,984 B2
5 6
(e.g. minimum values, minimum slopes, maximum slopes), proximate to the heart 12. Other examples of therapy systems
Zero crossings, threshold crossings, most negative slope, etc. may include three transvenous leads located as illustrated in
of a near or far-field EGM), onset of application of a pacing FIGS. 1-3. Still further, other therapy systems may include a
electrical stimulus, or the like. After a fiducial point is single lead that extends from the IMD 16 into the right atrium
selected, activation times are determined by measuring time 26 or the right ventricle 28, or two leads that extend into a
between the delivery of pacing stimulus using a pacing elec respective one of the right atrium 26 and the right ventricle 28.
trode and the appropriate fiducial point with the electrical FIG. 4 is a functional block diagram of one exemplary
activity sensed by a non-pacing electrode. The device deliv configuration of the IMD 16. As shown, the IMD 16 may
ering the pacing signal may include appropriate electronics to include a control module 81, a therapy delivery module 84
track and mark the timing of the pacing signal, which marked 10 (e.g., which may include a stimulation generator), a sensing
or tracked time may be used for purposes of determining local module 86, and a power source 90.
activation time and electrical dispersion as discussed above. The control module 81 may include a processor 80,
The device that delivers the pacing signal may be a device memory 82, and a telemetry module 88. The memory 82 may
configured for delivering CRT. include computer-readable instructions that, when executed,
As described in further detail with reference to FIG. 4, the 15 e.g., by the processor 80, cause the IMD 16 and/or the control
housing 60 may enclose a therapy delivery module that may module 81 to perform various functions attributed to the IMD
include a stimulation generator for generating cardiac pacing 16 and/or the control module 81 described herein. Further, the
pulses and defibrillation or cardioversion shocks, as well as a memory 82 may include any Volatile, non-volatile, magnetic,
sensing module for monitoring the patient's heart rhythm. optical, and/or electrical media, such as a random access
Cardiac pacing involves delivering electrical pacing pulses to memory (RAM), read-only memory (ROM), non-volatile
the patient’s heart, e.g., to maintain the patient’s heartbeat RAM (NVRAM), electrically-erasable programmable ROM
(e.g., to regulate a patient's heart beat, to improve and/or (EEPROM), flash memory, and/or any other digital media.
maintain a patient's hemodynamic efficiency, etc.). Cardiac The processor 80 of the control module 81 may include any
pacing involves delivering electrical pacing pulses ranging one or more of a microprocessor, a controller, a digital signal
from about 0.25 volts to about 8 volts and more preferably, 25 processor (DSP), an application specific integrated circuit
between 2-3 volts. (ASIC), a field-programmable gate array (FPGA), and/or
The leads 18, 20, 22 may also include elongated electrodes equivalent discrete or integrated logic circuitry. In some
62, 64, 66, respectively, which may take the form of a coil. examples, the processor 80 may include multiple compo
The IMD 16 may deliver defibrillation shocks to the heart 12 nents, such as any combination of one or more microproces
via any combination of the elongated electrodes 62, 64, 66 30 sors, one or more controllers, one or more DSPs, one or more
and the housing electrode 58. The electrodes 58, 62, 64, 66 ASICs, and/or one or more FPGAs, as well as other discrete
may also be used to deliver cardioversion pulses to the heart or integrated logic circuitry. The functions attributed to the
12. Further, the electrodes 62, 64, 66 may be fabricated from processor 80 herein may be embodied as software, firmware,
any Suitable electrically conductive material. Such as, but not hardware, or any combination thereof.
limited to, platinum, platinum alloy, and/or other materials 35 The control module 81 may control the therapy delivery
known to be usable in implantable defibrillation electrodes. module 84 to deliver therapy (e.g., electrical stimulation
Since electrodes 62, 64, 66 are not generally configured to therapy Such as pacing) to the heart 12 according to a selected
deliver pacing therapy, any of electrodes 62, 64, 66 may be one or more therapy programs, which may be stored in the
used to sense electrical activity during pacing therapy (e.g., memory 82. More specifically, the control module 81 (e.g.,
for use in determining activation times). In at least one 40 the processor 80) may control the therapy delivery module 84
embodiment, the LV elongated electrode 64 may be used to to deliver electrical stimulus Such as, e.g., pacing pulses with
sense electrical activity of a patient’s heart during the delivery the amplitudes, pulse widths, frequency, or electrode polari
of pacing therapy. Electrodes used to sense a response from ties specified by the selected one or more therapy programs
cardiac tissue are transmitted to an ND converter to convert (e.g., pacing therapy programs, pacing recovery programs,
the analog signal to a digital signal. The digital signal is then 45 capture management programs, etc.). As shown, the therapy
transmitted to the microprocessor 80. The microprocessor 80 delivery module 84 is electrically coupled to electrodes 40,
determines the level of response sensed at a particular elec 42, 44, 45, 46, 47, 48.50, 58, 62. 64, 66, e.g., via conductors
trode. of the respective lead 18, 20, 22, or, in the case of housing
The configuration of the exemplary therapy system 10 electrode 58, via an electrical conductor disposed within
illustrated in FIGS. 1-2 is merely one example. In other 50 housing 60 of IMD 16. Therapy delivery module 84 may be
examples, the therapy system may include epicardial leads configured to generate and deliver electrical stimulation
and/or patch electrodes instead of or in addition to the trans therapy Such as pacing therapy to the heart 12 using one or
venous leads 18, 20, 22 illustrated in FIG.1. Further, in one or more of the electrodes 40, 42, 44, 45, 46, 47, 48, 50, 58, 62,
more embodiments, the IMD 16 need not be implanted within 64, 66.
the patient 14. For example, the IMD 16 may deliver defibril 55 For example, therapy delivery module 84 may deliver pac
lation shocks and other therapies to the heart 12 via percuta ing stimulus (e.g., pacing pulses) via ring electrodes 40, 44.
neous leads that extend through the skin of the patient 14 to a 48 coupled to leads 18, 20, and 22, respectively, and/or helical
variety of positions within or outside of the heart 12. In one or tip electrodes 42, 46, and 50 of leads 18, 20, and 22, respec
more embodiments, the system 10 may utilize wireless pac tively. Further, for example, therapy delivery module 84 may
ing (e.g., using energy transmission to the intracardiac pacing 60 deliver defibrillation shocks to heart 12 via at least two of
component(s) via ultrasound, inductive coupling, RF, etc.) electrodes 58,62, 64, 66. In some examples, therapy delivery
and sensing cardiac activation using electrodes on the can/ module 84 may be configured to deliver pacing, cardiover
housing and/or on Subcutaneous leads. sion, or defibrillation stimulation in the form of electrical
In other examples of therapy systems that provide electri pulses. In other examples, therapy delivery module 84 may be
cal stimulation therapy to the heart 12, Such therapy systems 65 configured deliver one or more of these types of stimulation in
may include any suitable number of leads coupled to the IMD the form of other signals, such as sine waves, square waves,
16, and each of the leads may extend to any location within or and/or other Substantially continuous time signals.
US 8,929,984 B2
7 8
The IMD 16 may further include a switch module 85 and If IMD 16 is configured to generate and deliver pacing
the control module 81 (e.g., the processor 80) may use the pulses to the heart 12, the control module 81 may include a
switch module 85 to select, e.g., via a data/address bus, which pacer timing and control module, which may be embodied as
of the available electrodes are used to deliver therapy such as hardware, firmware, Software, or any combination thereof.
pacing pulses for pacing therapy, or which of the available The pacer timing and control module may include one or
electrodes are used for sensing. The switch module 85 may more dedicated hardware circuits, such as an ASIC, separate
include a Switch array, Switch matrix, multiplexer, or any from the processor 80. Such as a microprocessor, and/or a
other type of switching device suitable to selectively couple software module executed by a component of processor 80,
the sensing module 86 and/or the therapy delivery module 84 which may be a microprocessor or ASIC. The pacer timing
to one or more selected electrodes. More specifically, the 10 and control module may include programmable counters
therapy delivery module 84 may include a plurality of pacing which control the basic time intervals associated with DDD,
output circuits. Each pacing output circuit of the plurality of VVI, DVI, VDD, AAI, DDI, DDDR, VVIR, DVIR, VDDR,
pacing output circuits may be selectively coupled, e.g., using AAIR, DDIR and other modes of single and dual chamber
pacing. In the aforementioned pacing modes, "D' may indi
the Switch module 85, to one or more of the electrodes 40, 42, 15 cate dual chamber, “V” may indicate a ventricle, “I” may
44, 45, 46, 47, 48, 50, 58,62, 64, 66 (e.g., a pair of electrodes indicate inhibited pacing (e.g., no pacing), and “A” may indi
for delivery of therapy to a pacing vector). In other words, cate an atrium. The first letter in the pacing mode may indicate
each electrode can be selectively coupled to one of the pacing the chamber that is paced, the second letter may indicate the
output circuits of the therapy delivery module using the chamber in which an electrical signal is sensed, and the third
switching module 85. letter may indicate the chamber in which the response to
The sensing module 86 is coupled (e.g., electrically sensing is provided.
coupled) to sensing apparatus, which may include, among Intervals defined by the pacer timing and control module
additional sensing apparatus, the electrodes 40, 42, 44, 45.46. within control module 81 may include atrial and ventricular
47, 48, 50,58, 62, 64, 66 to monitor electrical activity of the pacing escape intervals, refractory periods during which
heart 12, e.g., electrocardiogram (ECG)/electrogram (EGM) 25 sensed P-waves and R-waves are ineffective to restart timing
signals, etc. The ECG is a record of the electrical activity of of the escape intervals, and/or the pulse widths of the pacing
the heart as the impulse travels from the atria through the pulses. As another example, the pacer timing and control
ventricles. The record is displayed in a waveform with three module may define a blanking period, and provide signals
distinct waves: P, QRS, and T. The ECG/EGM signals may be from sensing module 86 to blank one or more channels, e.g.,
used to monitor heart rate (HR), heart rate variability (HRV), 30 amplifiers, for a period during and after delivery of electrical
heart rate turbulence (HRT), deceleration/acceleration capac stimulation to the heart 12. The durations of these intervals
ity, deceleration sequence incidence, T-wave alternans may be determined in response to stored data in memory 82.
(TWA), P-wave to P-wave intervals (also referred to as the The pacertiming and control module of the control module 81
P-P intervals or A-A intervals), R-wave to R-wave intervals may also determine the amplitude of the cardiac pacing
(also referred to as the R-R intervals or V-V intervals), P-wave 35 pulses.
to QRS complex intervals (also referred to as the P-R inter During pacing, escape interval counters within the pacer
vals, A-V intervals, or P-Q intervals). QRS-complex mor timing/control module may be reset upon sensing of R-waves
phology, ST segment (i.e., the segment that connects the QRS and P-waves. Therapy delivery module 84 (e.g., including a
complex and the T-wave), T-wave changes, QT intervals, stimulation generator) may include one or more pacing out
electrical vectors, etc. 40 put circuits that are coupled, e.g., selectively by the Switch
The switch module 85 may also be used with the sensing module 85, to any combination of electrodes 40, 42, 44, 45.
module 86 to select which of the available electrodes are used, 46, 47, 48, 50, 58, 62, or 66 appropriate for delivery of a
e.g. to sense electrical activity of the patients heart. In some bipolar or unipolar pacing pulse to one of the chambers of
examples, the control module 81 may select the electrodes heart 12. The control module 81 may reset the escape interval
that function as sensing electrodes via the Switch module 45 counters upon the generation of pacing pulses by therapy
within the sensing module 86, e.g., by providing signals via a delivery module 84, and thereby control the basic timing of
data/address bus. In some examples, the sensing module 86 cardiac pacing functions, including anti-tachyarrhythmia
may include one or more sensing channels, each of which pacing.
may include an amplifier. In some examples, the control module 81 may operate as an
In some examples, sensing module 86 includes a channel 50 interrupt driven device, and may be responsive to interrupts
that includes an amplifier with a relatively wider pass band from pacer timing and control module, where the interrupts
than the R-wave or P-wave amplifiers. Signals from the may correspond to the occurrences of sensed P-waves and
selected sensing electrodes that are selected for coupling to R-waves and the generation of cardiac pacing pulses. Any
this wide-band amplifier may be provided to a multiplexer, necessary mathematical calculations may be performed by
and thereafter converted to multi-bit digital signals by an 55 the processor 80 and any updating of the values or intervals
analog-to-digital converter (ND) for storage in memory 82 as controlled by the pacer timing and control module may take
an electrogram (EGM). In some examples, the storage of such place following such interrupts. A portion of memory 82 may
EGMs in memory 82 may be under the control of a direct be configured as a plurality of recirculating buffers, capable
memory access circuit. The control module 81 (e.g., using the of holding series of measured intervals, which may be ana
processor 80) may employ digital signal analysis techniques 60 lyzed by, e.g., the processor 80 in response to the occurrence
to characterize the digitized signals stored in memory 82 to of a pace or sense interrupt to determine whether the patients
detect and classify the patients heart rhythm from the elec heart 12 is presently exhibiting atrial or ventricular tach
trical signals. For example, the processor 80 may be config yarrhythmia.
ured to measure activation times of cardiac tissue using The telemetry module 88 of the control module 81 may
EGMs from one or more electrodes in contact, or in proxim 65 include any suitable hardware, firmware, Software, or any
ity, with cardiac tissue by employing any of the numerous combination thereof for communicating with another device,
signal processing methodologies known in the art. Such as the programmer 24 as described herein with respect to
US 8,929,984 B2
10
FIG.1. For example, under the control of the processor 80, the nals to digital signals. Digital signals are then transmitted to
telemetry module 88 may receive downlink telemetry from the microprocessor 80 so that signals can be measured and
and send uplink telemetry to the programmer 24 with the aid then stored into memory 82 at operation 104.
of an antenna, which may be internal and/or external. The At block 106, after obtaining the electrical activation times
processor 80 may provide the data to be uplinked to the (e.g. determined with respect to the timing of the earliest
programmer 24 and the control signals for the telemetry cir Ventricular pacing or any other suitable means) at non-pacing
cuit within the telemetry module 88, e.g., via an address/data electrodes, the microprocessor 80 determines the weighted
bus. In some examples, the telemetry module 88 may provide electrical dyssynchrony index for the first fusion pacing con
received data to the processor 80 via a multiplexer. In at least figuration. Electrical dyssynchony or cardiac dyssynchrony
one embodiment, the telemetry module 88 may be configured 10 involves improperly timed electrical activation of one or more
to transmit an alarm, or alert, if the pacing therapy becomes different parts of the heart.
ineffective or less effective. The fusion index of LV electrical dyssynchrony IFI (, A)
The various components of the IMD 16 are further coupled can be computed for each pacing electrode j from a linear
to a power source 90, which may include a rechargeable or combination of electrical activation times (LVAT(i., A)) at
non-rechargeable battery. A non-rechargeable battery may be 15 each non-pacing LV electrode denoted by i and RV electrical
selected to last for several years, while a rechargeable battery activation times RVAT(i., A) at each RV electrodei. “A” of FI
may be inductively charged from an external device, e.g., on (A) refers to the atrioventricular delay between atrial sense
a daily or weekly basis. (or pace) and the Ventricular pacing pulse. A determination of
After the LV lead 20 has been properly positioned on or FI for each LV electrode may be made initially at a nominal
near the LV tissue, schematically shown in FIG.3, and the RV value of A such as 50 ms.
lead is in position, a variety of fusion pacing configurations FI is determined by a weighted linear combination of elec
(e.g. RV only pacing configuration, LV only pacing etc.) can trical activation times in which individual weights are deter
be tested. Data generated from each pacing configuration can mined depending upon the lead-geometry and the inter-elec
be useful in determining the optimal LV electrode from which trode spacing on the lead. In particular, FI is weighted by a
to pace the LV or the optimal RV electrode to pace the RV. 25 suitable factor w(i, j) that is based on the distance of the
Each fusion pacing configuration employs a different LV non-pacing electrode (designated 'i') from the pacing elec
electrode (e.g. LV1, LV2, LV3, and LV4 etc.) or a different RV trode (designated ''). Accordingly, the equation for calcu
electrode for pacing. lating a weighted FI is as follows:
Exemplary methods and/or devices described herein evalu
ate the effectiveness of cardiac resynchronization based on 30
metrics of electrical dyssynchrony derived from the measured
cardiac electrical activation times for each fusion pacing con where “n” is the total number of LV electrodes and m is the
figuration employing a different LV electrode. FIGS. 5-7 flow total number of RV electrodes.
diagrams present different exemplary methods for selecting Only valid FI are used to determine an optimal LV elec
an optimal LV electrode or RV electrode. 35 trode from which to pace. Before the electrode evaluations are
Exemplary method 100, depicted in FIG. 5, evaluates a performed, the left ventricular capture management routine
fusion pacing configuration Such as LV only pacing in order may be evoked to determine the minimum thresholds
to determine which LV electrode on lead 20 is optimal for required for left ventricular capture for each LV pacing vec
pacing the LV. Each of the available fusion pacing configu tor. The evaluations of FI are performed while pacing the LV
rations, based on the selected LV only pacing electrode, is 40 at outputs with adequate margins V) above the minimum
serially tested and evaluated by programmer 24 as to its threshold determined by the left ventricular capture manage
effectiveness based on the metrics of electrical dyssynchrony. ment routine for each LV pacing vector.
The optimal fusion pacing configuration is selected based on Additionally, automatic LV capture detection is turned on
one or more of these metrics. While the methods are described to ensure that the pacing pulse delivered captures each ven
relative to LV only pacing, skilled artisans appreciate that the 45 tricle. FI is not computed for instances where the pacing pulse
same principles can be applied to RV only pacing. At block does not capture the LV. Capture detection can be verified by
102, the programmer 24 switches one of the LV electrodes 44. determining the amplitude of an evoked response at the pac
45, 46, 47 to a pacing mode while the other LV electrodes ing electrode within a short duration of time after delivery of
remain in the sensing mode. The LV electrode that is selected the pace. More particularly, capture detection can be verified
for pacing the LV is designated as the j-th LV electrode. The 50 by observing an initial negative deflection within 20-60 ms
first LV electrode out of the plurality of LV electrodes to pace after the pace delivery, on the EGM viewed between the
the LV is referred to in the claims as the first LV electrode. The pacing electrode and an indifferent electrode like the device
programmer 24 includes a pulse generator that generates case or an RV coil electrode. Exemplary capture can be indi
pacing pulses (e.g. 2-3 volts amplitude) that are delivered cated by an amplitude greater than 0.5 mV.
through the pacing LV electrode to the LV. In one or more 55 The FI may be calculated over multiple (N) beats, where N
embodiments, the first LV electrode is paced at paced A-V can be any number between 5 to 10, during pacing from a
delays (PAV) or SAV at least 60 ms shorter than the intrinsic selected LV electrode to ensure that measurements are con
delay. Intrinsic A-V delay is determined through the formula sistent and repeatable. A coefficient of variation which is
below: provided by the standard deviation of FI divided by mean FI
60 index over N number of beats multiplied by 100 (to express as
Ventricular sensed event time(Vs)-Atrial-sensed a percentage) may be computed to measure the amount of
event time(As)Ventricular sensed event time variability in FI for a given LV pacing vector. If the coefficient
(VS)-Atrial-paced event time(Ap).
of variation is less than a certain percentage threshold (which
The IMD is configured to determine the timing and interval can be any number from 5% to 20%), the FI measurements are
between these events. The electrogram signals at each of the 65 considered to be valid and the mean or the median of the FI
non-pacing LV electrodes as well as the RV electrode are values may be taken as the representative measure of FI for
transmitted to an A/D converter that converts the analog sig that particular LV pacing electrode.
US 8,929,984 B2
11 12
FI is typically computed with theatrioventricular delay (A) Since LV2 and LV3 are substantially close, the AT of LV2 and
set at a constant value when evaluating each LV electrode. For LV3 are averaged together. The AT associated with LV4 is
example, the A-V delay can be set at a preselected value (e.g. multiplied by a constant number w(i,j). W(i,j) is a weighting
50 to about 140 ms etc.) for each fusion pacing configuration factor that depends on the distance from LV 1 to LV4 as
where the pacing cathode is LV1, LV2,.... LVn. In another 5 compared to the distance between LV electrodes (2.3) and
embodiment, the pre-selected AV delay (A) may be deter LV1. Skilled artisans will appreciate that although the RV
mined from the intrinsic AV delay (iAV) by the following bipolar lead consists of two electrodes (e.g. RV tip and RV
scheme: ring), only RVAT (1) is represented in the equation since the
RV electrodes are closely spaced to one another. Closely
ifi AV-60 mss.80 ms, 10 spaced electrodes measure about the same activation time.
A=iAV-60 ms
However, if the RV electrodes were not closely spaced, the
equation could be modified to include the second RV elec
else
trode. In one or more embodiments, since only LV electrodes
are selected, a uniform weighting factor (e.g. “1) can be
15 assigned to each RV electrode(s).
A=80 ms
In this example, w(4,1) is /2 since the distance from LV 1 to
Assume A=50 ms while evaluating each of the LV elec LV4 is twice as long as the distance from electrodes (2.3) to
trodes as is discussed in greater detail below. Additionally, LV1. W(i,j) can be adjusted depending upon the LV medical
skilled artisans appreciate that the activation time for the electrical lead and the spacing used between the plurality of
pacing electrode which captures the tissue may be 0 or may be electrodes thereon.
skipped. Evaluation of FI and activation propagation is also per
After the FI has been determined for the first LV electrode, formed while pacing from other LV electrodes 2, 3 and 4 in
the programmer 24 automatically selects a second fusion the quadripolar lead 20 during fusion pacing. The equation
pacing configuration in which a second LV electrode paces for calculating weighted FI at LV2 or LV3 is as follows:
the LV at operation 108. The programmer 24 causes the pulse 25
generator to generate pacing pulses (e.g. 2-3 volts amplitude)
through the second LV electrode to the LV. The non-pacing
LV electrodes record the electrical response (e.g. electro
grams.) from the LV tissue. The electrogram signals are trans
30
mitted to an A/D converter that converts the analog signals to Since the spacing between LV2 and LV3 is less than 2 mm,
digital signals. Digital signals are then transmitted to the and LV1 and LV4 are about equidistant from LV2 and LV3,
microprocessor 80 so that electrogram signals can be mea the activation times are weighted equally while computing FI
Sured, activation times can be computed and then stored into for LV2 and LV3.
memory 82 at operation 110. After obtaining the electrical The equation for calculating a weighted FI at LV4 is as fol
activation times at non-pacing electrodes, the microprocessor 35
lows:
80 determines a weighted fusion index associated with the
second LV pacing electrode at operation 112. Skilled artisans
appreciate that blocks 108-112 are repeated for all of the
remaining LV electrodes (e.g. LV3, LV4 etc.) at the distal end
of lead 20. 40 After weighted FI calculations are performed, the optimal LV
Once two or more valid FI have been calculated, electrode electrode is selected at block 114.
elimination rules can be applied to the FI to eliminate LV FIGS. 6A-6B provide exemplary methods in which an
electrodes in order to determine the optimal LV electrode at optimal electrode is selected from a set of electrodes to per
operation 114. The FI for each electrode can be determined form fusion pacing. FIG. 6A, for example, shows a method
for the set of electrodes before application of the set of rules. 45 200 in which a process of electrode elimination is used to
Alternatively, the rules can be applied after determining a FI determine the optimal LV electrode from the plurality LV
for any two electrodes and once one of the two electrodes is electrodes (e.g. LV1, LV2, LV3, and LV4). The process of
eliminated, a FI is calculated for yet another electrode to be elimination employs two different types of electrode com
compared to the FI of the remaining electrode. parisons that are used to eliminate an electrode from each pair
Examples of the manner in which FI are calculated are 50 of electrodes until the sole remaining electrode is deemed to
presented below as to the set of LV electrodes (i.e. left ven be the optimal LV electrode. The process of eliminating an
tricle electrode 1 (LV1)44, left ventricle electrode 2 (LV2)45, electrode begins at block 202 in which activation times are
left ventricle electrode 3 (LV3)46, and left ventricle 4 (LV4) measured at the LV electrodes (e.g. LV1, LV2, LV3, and LV4)
47 etc.) shown on the LV medical electrical lead 20 and a RV during baseline rhythm which may constitute RV only pacing
bipolar lead; however, it is appreciated that teachings pre 55 or intrinsic rhythm. RV only pacing occurs when the pulse
sented herein can be applied to two or more LV electrodes on generator from the programmer 24 delivers electrical stimu
a medical electrical lead. lation (i.e. pacing pulses) through an RV electrode to the RV
While pacing from LV1 during fusion pacing and A=50 and none of the LV electrodes are used to pace the LV. Sensing
ms, the FI is computed below. For this nominal value of A, the the activation times at all of the LV electrodes (e.g. LV1, LV2.
FI for the fusion pacing with the j-th LV electrode and the 60 LV3, and LV4) during RV only pacing or during intrinsic
activation time at the i-th LV electrode during such fusion rhythm can be performed any time after LV lead 20 has been
pacing are represented by FI () and AT(i) respectively. The placed near and/or on LV tissue.
weighted FI for pacing at LV1 can be rewritten as follows: At block 204, the activation times associated with each of
the LV electrodes are stored into the memory 82. At block
65 206, variables threshold (T) level, integer (I), and total num
ber of electrodes (N) (e.g. N=4 on the distal end of lead 20) are
initialized, set, and stored into memory 82. Threshold T can
US 8,929,984 B2
13 14
be predetermined and input into the programmer 24 by the eliminated based upon additional or other criteria Such as
user before evaluating each LV electrode (e.g. LV1, LV2. lower capture threshold, higher impedance (i.e., reduced
LV3, LV4). Preferably, T equals 15 ms or less. A value of T energy required to pace), or absence of phrenic stimulation
equal to 15 ms or less can be typical of a left bundle branch could be considered (by the user) to select the best electrode
block (LBBB) patient with a QRS of 150 ms. Additionally, T of two electrodes that have equivalent ATs.
equal to 15 ms or less is typically about a 10% time of total At block 226, one electrode (i.e. LV1) is eliminated. I is
ventricular activation. In one or more other embodiments, T incremented by 1 at block 230. At block 232, a determination
equals 10 ms or less. is made as to whether I=N, which essentially determines
I and Nare used in a counting loop (i.e. blocks 206, 230 and whether FI data has yet to be retrieved.
232) that ensures that data for each LV electrode (e.g. LV1, 10 Returning to block 224, if one baseline AT is not less than
LV2, LV3, and LV4) are analyzed before the optimal LV another baseline AT, the NO path can be followed to block
electrode is selected. I=1 since FI is determined for only one 228 in which another electrode (i.e. LV2) is eliminated. The
LV electrode at block 208 by processor 80 and stored in counting loop at block 232 is then used to determine whether
memory 82. The data for determining the FI for one LV any additional FI data must be processed. At block 232, once
electrode can be selected from any one of the LV electrodes 15 I=N, no additional FI data needs to be processed. Therefore,
(e.g. LV1, LV2, LV3, and LV4). At block 210, data for another the YES path can be followed to block 234. The optimal
LV electrode is retrieved from memory 82 by processor 80. electrode is then designated as the LV electrode that remains
Determining the FI for another LV electrode means any other or has not been eliminated. The optimal LV electrode is set to
LV electrode data not previously analyzed. For example, if pace the LV automatically by programmer 24 or manually by
the FIfor one LV electrode at block 208 is data related to LV1, the user.
then data for another LV electrode can be related to LV2, LV3, Examples, presented below, show the electrode elimina
or LV4. For the sake of illustration, assume that the data for tion process used to select an optimal LV electrode. In these
another LV electrode is associated with LV2. Therefore, the examples, assumptions are made. For blocks 222, and 224,
FI for LV2 is calculated by processor 80 and stored in memory activation times for the LV electrodes are performed during
82. 25 RV only pacing or during intrinsic rhythm. In contrast, the FI
At block 212, the difference in magnitude between one FI data was generated using the fusion pacing configurations, as
data and another FI data is determined. For example, the FI previously described herein. Additionally, a quadripolar LV
data for one electrode (i.e. LV2) is subtracted from FI data for lead 20 is used that includes four LV electrodes, LV1, LV2,
another electrode (i.e. LV1). At block 214, the difference LV3, and LV4; however, skilled artisans appreciate that other
between one FI data (i.e. FI 1) and another FI data (i.e. FI2) 30 embodiments could use two or more LV electrodes on a lead
is compared to a threshold level T (also referred to as delta T 20 such as two or more electrodes on one lead and two or more
or AT). At block 216, if the difference is not less than T, then electrodes on another lead. Each example will be described
the NO path can be followed to block 218. At operation 218, relative to FIG. 6A.
whichever electrode is associated with a larger FI is automati In the first example, assume that the order of activation
cally eliminated from consideration as a potential optimal LV 35 times during RV only pacing or intrinsic rhythm is AT(LV4)>
electrode irrespective of the eliminated electrode's activation AT(LV1)>AT(LV2)>AT(LV3) with LV4 being the latest acti
time obtained during intrinsic rhythm or RV only pacing. vation time and LV3 being the earliest activation time.
The counting loop increases the variable I by one at block Assume also that the values of FI, determined from the fusion
230. At block 232, a determination is made as to whether I=N. configurations previously discussed are as follows: FI(1)=50
Since after the first pass of the counting loop I-2 and N=4, the 40 ms, FI(2)=55 ms. FI(3)=58 ms, and FI(4)–74 ms. Other
NO path transfers control to block 210 to retrieve FI data for assumptions include A is a constant (e.g., 50 ms) and a pre
yet another LV electrode. FI is then calculated for yet another determined thresholdT of 15 ms is used to analyze the FI data.
LV electrode (e.g. LV3) and then stored into memory 82. Processor 80 retrieves FI data such as FI(1) and FI(2) at
Skilled artisans appreciate that after an electrode is elimi blocks 208, 210, respectively.
nated, at either block 218 or 228, and FI data for another 45 At block 212, the difference in magnitude between FI(1)
electrode is retrieved, a Swapping operation may be per and FI(2) is calculated as follows:
formed. For example, if data for LV2 is initially designated as
"another LV electrode' and LV2 is eliminated, then data for FI(2)-FI(1)=55 ms-50 ms=5 ms
LV3 is swapped for the FI data for LV2 and the FIdata for LV3 At block 214, the difference in FI(1) and FI(2) is compared
is now stored in the register for “another FI data at block 210. 50 to the threshold T. At block 216, a determination is made as to
The electrode pair comparisons are then between LV1 and whether the difference in FI(1) and FI(2) is less than the
LV3 and so on. predetermined threshold of 15 ms. Since the difference (i.e. 5
Returning to block 216, if the difference in FI value is less ms) in FI(1) and FI(2) is less than T, the YES path is followed
than T, the YES path transfers control to block 222. At block to block 222 in which the AT values (i.e. AT(1), AT(2)) are
222, the baseline (intrinsic rhythm or RV only pacing) AT data 55 compared. In one or more embodiments, the compare func
for one electrode (i.e. LV1) is compared to the baseline AT tion can also include sorting the activation times in ascending
data for another electrode (i.e. LV3). The baseline AT data is order or descending order.
preferably obtained during intrinsic rhythm or RV only pac At block 223, if one baseline AT is equal to the other
ing. Comparing one baseline AT to another baseline AT can baseline AT, then both electrodes are retained in a preference
involve a sorting function that places the data in ascending 60 list. One of the two electrodes is eliminated based upon the
order or descending order. At block 224, a determination is previously described criteria at block 225. If one AT does not
made as to whether one baseline AT is less than another equal another AT, the NO path goes to block 224.
baseline AT. If one baseline AT is less than another baseline At block 224, a determination is made as to whether one AT
AT, the YES path can be followed to block 226. (i.e. AT1) is less than another AT (i.e. AT2). As is known from
Returning to block 223, if one baseline AT is equal to the 65 the given facts, the activation time for one AT (i.e. AT1) is
other baseline AT, then both electrodes are retained in a pref greater than another AT (i.e. AT2). The NO path can be fol
erence list. At block 225, one of the two electrodes can be lowed to block 228, which causes the elimination of another
US 8,929,984 B2
15 16
electrode (i.e. LV2). The variable I is increased by 1 at block At block 212, the difference between FI(1) and FI(3) is
230. At block 232, a determination is made as to whether I=N. calculated as follows:
Since I-2 and N=4, I does not equal N. The NO path returns
to block 210 for processor 80 to retrieve from memory 82 FI(1)-FI(3)=30 ms-25 ms=5 ms
another FI value for another LV electrode (e.g. LV3). The difference in FI values of LV1 and LV3 is 5 ms which
The FI value of LV3 is then Subtracted from the FI value for is less than the threshold value of 15 ms at block 216. The YES
LV1 at block 212, as shown below. path can be followed to block 222 which compares AT(1) to
AT(3). Since AT(3) is greater than AT(1), the electrode LV3 is
FI(3)-FI(1)=58 ms-50 ms=8 ms eliminated at block 228 based on its earlier activation time
10
The difference in magnitude (i.e. 8 ms) between FI(1) and compared to the electrode LV1. Again, I is incremented by 1
at block 230 and another determination is made as to whether
FI(3) is less than the predetermined threshold of 15 ms at
block 216. The YES path can be followed to block 222. At I=N at block 232. Since I-3, I does not equal N. Therefore,
block 222, the activation times between AT(1) and AT(3) are another FI data such as FI(4) is retrieved from memory 82.
compared to each other. As previously stated, AT(1) is greater 15
LV4 can then be compared with the electrode LV1. The
difference in FI values of electrode LV1 and the electrode
than AT(3). LV3 is then eliminated based on its earlier acti LV4 is 2 ms. At block 224, one AT is found to be less than
vation time compared to LV1 at block 228.
At block 230, the variable I is again increased by 1 which another AT. The electrode LV1 is eliminated due to AT(LV1)
causes I-3. A determination is made as to whether I=N at having an earlier activation time compared to AT(4). Again, I
block 232. Since I does not equal N, the NO path is followed is incremented by 1 causing I-4. Since I=N at block 232, the
to block 210. The FI data for the next electrode, FI(4), is then optimal electrode is LV4. LV4 is chosen as the final or optimal
then retrieved at block 210. electrode from which to pace the LV since LV4 was not
eliminated.
FI(1) data, associated with LV1, is subtracted from FI(4) at A third example is presented in which activation times
block 212. At block 214, the difference in FI values is 24ms,
which is greater than the pre-determined threshold of 15 ms. during RV only pacing or intrinsic rhythm of the four elec
25
The NO path can be followed to block 218 in which the trodes are such that AT(LV4)>AT(LV1)>AT(LV2)>AT(LV3).
electrode to be eliminated is associated with the larger FI data. Additionally, the FI values generated from the fusion pacing
The electrode with the larger FI, i.e. LV4, is eliminated with configurations are FI(1)=60 ms, FI(2)-40 ms, FI(3)-38 ms.
out any comparison being performed between the activation and FI(4)=62 ms. Referring to FIG. 6, processor 80 retrieves
times of LV1 and LV4. 30
FI(1) data and FI(2) data from memory 82 at blocks 208,210,
At block 230. I is again incremented by 1 causing I-4. At respectively. At block 212, the difference in FI values can be
block 232, a determination is made as to whether I=N. Since calculated by the following:
I=4 and N=4, then I=N. The YES path can be followed to FI(1)-FI(2)=60 ms-40 ms=20 ms
block 234, which designates the optimal electrode is LV1
since LV1 is the last remaining electrode that was not elimi 35 At block 216, since the difference in FI values associated
nated in the exhaustive electrode-pair comparisons. LV1 is with LV1 and LV2 is 20 ms which exceeds the threshold of 15
then selected as the final electrode for delivering CRT. ms, the NO path can be followed to block 218 in which the
A second example shows how the selection is made when electrode with the higher FI value, i.e. electrode LV1 is elimi
FI values of all electrodes are almost equivalent or similar. For nated. I is incremented by 1 at block 230. A determination is
example, assume the order of activation times during intrinsic 40 then made as to whether I=N at block 232. Since I=2, and N
rhythm (or RV only pacing) are such that AT(LV4)>AT equals 4, the NO path returns to block 210 to retrieve FI(3)
(LV1)>AT(LV2)>AT(LV3). LV4 is associated with the latest data.
activation time and LV3 is associated with the earliest activa At block 212, the difference in FI values can be calculated
tion time. From the fusion pacing configurations, the FI val as follows:
ues were determined such that FI(1)–30 ms, FI(2)-33 ms. 45
FI(3)=25ms, and FI(4)=28 ms. Referring to FIG. 6A-B, FI
data is retrieved for one LV electrode such as FI(1) at block Since the difference in magnitude between FI(2) and FI(3)
208. At block 210, FI data is retrieved for another LV elec is 2ms, which is less than the threshold of 15 ms, the YES path
trode such as FI(2) at block 212. At block 212, the difference can be followed to block 222. At block 222, one AT (i.e. LV3)
in FI values is calculated follows: 50 is compared to another AT (i.e. LV2). LV3 is associated with
a AT that is less than the AT for LV2. At block 224, a deter
FI(2)-FI(1)=33 ms-30 ms=3 ms mination is made as to whether one AT (i.e. LV3) is less than
At block 214, the difference in FI(1) and FI(2) is compared another AT (i.e. LV2). At block 226, electrode LV3 is elimi
to predetermined threshold of 15 ms. As shown above, the nated.
difference in FI(1) and FI(2) is only 3 ms which is less than the 55 Again, I is incremented by 1 at block 230. Therefore, I=3.
predetermined threshold of 15 ms. At block 216, a determi At block 232, I does not equal N since I=3 and N=4; therefore,
nation is made as to whether the difference in FI values is less at block 210, FI(4) is retrieved from memory 82.
than the threshold. Since the difference is 3 ms is less than 15 At block 212, the difference between FI(4) and FI(2) can be
ms, the YES path can be followed to block 222 in which one shown as follows:
AT (i.e. AT(1)) is compared to another AT (i.e. AT(2)). From 60
FI(4)-FI(2)=62 ms-40 ms=22 ms
the comparison, it was determined that AT(1) is greater than
AT(2). At block 224, a NO path can be followed to block 228 Since the difference in their FI values is 22 ms, well above
that eliminates another electrode (i.e. LV2). At block 230. I is the threshold of 15 ms, the NO path can be followed to block
incremented by I causing I-2. At block 232, a determination 218. At block 218, the electrode associated with the larger FI
is made as to whether I=N. Since I=2 and N=4, I does not 65 value is eliminated. Since FI(4) is larger (i.e. 62 ms) than
equal N. Therefore, the NO path returns to block 210 in which FI(2) (i.e. 40ms), LV4 is eliminated. I is again incremented by
another FI data (i.e. FI3) is retrieved from memory 82. 1 at block 230 thereby causing I to be equal to 4. At block 232,
US 8,929,984 B2
17 18
I=N; therefore, the YES path can be followed to block 234.
The optimal electrode is LV2. LV2 is then used to pace the LV.
The method embodied in FIG. 6B is the same as FIG. 6A
except block 225 is replaced by block 227. As previously
described relative to block 223, if one baseline AT is equal to
the other baseline AT, then both electrodes are retained in a Skilled artisans will appreciate that although the RV bipolar
preference list. One of the two electrodes can be eliminated lead consists of two electrodes (e.g. RV tip and RV ring), only
based upon additional or other criteria. For example, the RVAT (1A) is represented in the equation since the RV elec
pacing pulse can be automatically adjusted (e.g. increased or trodes are closely spaced to one another. Closely spaced
decreased) at block 227. Equivalent electrodes are re-evalu 10 electrodes measure about the same activation time. However,
ated under method 200 by returning to block 202 using the if the RV electrodes were not closely spaced, the equation
new pacing criteria to determine whether a difference exists could be modified to include the second RV electrode.
between the two electrodes. For example, the pacing pulse To better understand the relationship between LVAT and
can be increased by 0.25 volts, 0.5 volts, 0.75 volts and so on. RVAT and how these times are calculated, it may be useful to
After rechecking the electrodes under method 200 using the 15 examine a ventricular electrogram, which shows changes in
increased pacing pulse, more than likely, a difference will electrical potential at the corresponding LV and RV elec
exist between the two electrodes and the electrode that under trodes (e.g. LV electrode-can, LV electrode-RV coil, RV elec
performs is eliminated. If not, the pacing criteria can again be trode-can, RV tip-RV ring etc.). FIG. 8, for example, shows an
modified and the electrodes rechecked under method 200. atrial event is sensed by the right atrial electrodes. Theatrial
The pacing criteria can be continuously adjusted and the event may be used as a timing marker or timing reference. The
electrodes evaluated under method 200 until a difference window, used to compute activation times for the depolariza
exists between the electrodes and one of the electrodes can be tion signals, can be defined as extending from theatrial event
eliminated. If one AT does not equal another AT, the NO path and ending after expiration of a certain time period (e.g. 400
goes to block 224. ms window timed from theatrial event). All computed acti
A set of LV electrode elimination rules can be summarized 25 vation times could be measured from the timing of the atrial
below which can be applied to scenarios in whichananatomic event.
block is present or not present. An anatomic block is a differ Referring to FIG. 8, after theatrial event is sensed by one of
ence between two AT that is greater than a threshold T. One the electrodes a ventricular pacing stimulus is then delivered
LV electrode elimination rule is that when all electrodes have through LV 1 to the left ventricle at an atrioventricular delay
equivalent FI values, the LV electrode with the latest activa 30 (A) while no pacing stimulus is delivered to the RV. The upper
tion during RV only pacing or intrinsic rhythm is selected for panel of FIG. 8 shows a RV far-field electrogram (RV ring
final CRT therapy. However, if one LV electrode is associated Can). The lower panel of FIG. 8 shows a far-field LV electro
with a significantly higher FI compared to another LV elec gram which is the electrical activity measured relative to the
trode (i.e. a difference exceeding the predetermined thresh electrode that is the greatest distance away from the pacing
old), the electrode with the higher FI is eliminated as a pos 35 electrode (i.e. measured from LV4-RV coil electrodes). Com
sible choice, irrespective of the activation times during putation of activation times from far-field electrograms
intrinsic rhythm or RV only pacing. involves determining the time corresponding to the steepest
In one or more embodiments, once an optimal LV electrode negative slope of the electrogram during the depolarization
is chosen, an optimal delay Such as A-V delay (A) can be cycle. These times for the RV and LV far-field electrograms
determined through exemplary method 300 presented in flow 40 are indicated by t-RV and t-LV on the upper and lower panels
diagram of FIG.7, respectively. In one or more embodiments, respectively. Once the appropriate timing on the waveform is
A-V delay optimization occurs in a similar manner as that determined, all activation times may be referenced to the
which was performed to select the optimal LV electrode from timing of the delivery of the Ventricular pace. Consequently,
which to pace. In one or more embodiments, A-V delay the right-ventricular activation time is RVAT(1A)=tRV-A
optimization can be performed through a use of a weighted 45 and LV activation time at electrode 4 is LVAT(1A)=tLV4-A
sum of activation times for various A-V delays. The A-V where tRV and tLV4 are the times corresponding to the most
delay that results in the lowest electrical dyssynchrony is negative slope with respect to each signal within the defined
selected and programmed into the programmer 24. window (i.e. 400 ms window). As shown in FIG. 8, tRV and
In order to determine the optimal A-V delay for fusion tLV4 extend from the beginning of the atrial event to time
pacing Such as LV only fusion pacing, a FI must be calculated 50 corresponding to the timing steepest slope of the correspond
for at least two or more A-V delays. FIC, A) represents the ing farfield electrograms. The window, as previously
electrical dyssynchrony during fusion pacing LV electrodej. described, starts with an atrial sensed event oran atrial paced
'A' represents an A-V delay and LVAT(i., A) and RVAT(i, A) event.
represents activation time at LV electrode i during fusion Though this specific example describes determination of
pacing and activation time at RV electrode i respectively. The 55 activation times from far-field electrograms, the same can be
FI equation is as follows: done from near-field electrograms. For near-field electro
FI(i.A)=X, 'w-(ii)|LVAT(i,Al)+X, 'w-(ii) IRVAT
grams (e.g. RV tip-RV ring), computation of activation time
(i,Al) involves determining the timing of the maximum peak or the
where “n” is the total number of LV electrodes and m is the
minimum valley. In case of a biphasic near-field waveform,
60 the point of Zero-crossing may be also taken as the corre
total number of RV electrodes used in computation of the sponding activation time. As in the example, activation times
fusion index. For a standard CRT implantable device involv on different electrodes may then be referenced with a com
ing a single bipolar RV lead and a quadripolar LV lead, fusion mon timing fiducial or marker, like the timing of the atrial
indices during pacing from each of the LV electrodes may be event).
computed as follows for a given A-V delay A: 65 As previously stated, only valid FI are used to determine an
optimal LV electrode from which to pace. Data is omitted
when the pacing stimulus delivered to LV fails to capture
US 8,929,984 B2
19 20
because of insufficient energy. In an alternative embodiment, FI is then calculated for the third A-V delay. After the third FI
LV pacing is delivered at maximum energy to prevent the is calculated, the programmer 24 automatically calculates up
scenario of failure to capture because of insufficient energy. to N number of A-V delays. Typically, the programmer 24
The programmer 24 can automatically choose A-V delays automatically tests N (e.g. N can be 12-20 etc.) number of
ranging from a lowest Value of 40 ms to a highest Value of 260 5 sensed A-V delays and M number of paced A-V delays for a
ms, in increments of 5, 10, 15 or 20 ms for atrial sensing. The resting cycle-length (e.g. time (ms) between two events such
same values are also selected during atrial pacing. as successive atrial events). Typically Nequals M, although
After selecting the A-V delays, the programmer 24 causes skilled artisans will understand that N does not have to be
the pulse generator to generate pacing pulses (e.g. ranging equal to M since determining FI values for sensed A-V delays
E. by'95. s by 8 SR E.E. Estably " programmer
etween 2-3 volts) that are delivere Ougn une opuma
is a different 24
operation than paced A-V delayS. Generally, the
tests less than 100 A-V delays. In one or more
electrode to the LV. The physiological response to the pacing other embodiments, programmer 24 can automatically test 20
1 be ob d.
pulses
Aftercan be observe
measuring the electrical activation times at non-pac or less A-V delays. In yet another embodiment, programmer
ing electrodes at operation 304, the microprocessor 80 deter- 15 24 can automatically test 10 or less A-V delays. Negative A-V
mines fusion index for the first A-V delay using the FI equa delays are not tested because pre-excitation of Ventricles
tion above associated with the optimal LV pacing electrode at before the atrial activation is not hemodynamically optimal.
operation306. For example, the weighted sum equation could Table 1, presented below, provides an example of FI results
take into account the physical spacing, as previously dis- for SA-V delays that ranges from a short delay (i.e. 40 ms) to
cussed, between the LV electrodes on the LV lead 20. For 20 along delay (i.e. 260 ms). Each A-V delay is automatically
example, the electrical dyssynchrony metric for fusion pacing separated by predetermined time increments (i.e. 20 ms)
from LV1 electrode at a A-V delay of 40 ms can be expressed although other Suitable time incremental values (e.g. 5 ms, 10
as follows: ms, 15 ms etc.) can also be used. Fusion pacing from LV or
RV electrodes performed using a particular A-V delay, while
FI(140)=AT(140)+AT(2,340)+AT(4.40)/2
AT(2,340)=AT(2,40)+AT(3,40)/2
wherein as maintaining the V-V delay at a constant or fixed nominal
value allows exemplary data to be generated for Table 1. The
Since LV2 and LV3 are substantially close, the AT of LV2 A-V delay that provides a minimum FI is selected as an
and LV3 are averaged together. The AT associated with LV4 optimal A-V delay. In this example, the optimal A-V delay is
is multiplied by a constant number W.W is the distance from 120 ms that corresponds to a minimum FI.
TABLE 1.
FI for a range of Sensed A-V (SA-V) delays at resting heart rate
SA-V delay (ms
40 60 80 100 120 140 160 180 200 220 240 260
FI(ms) 39 39 32 28 25 29 36 41 45 46 46 46
LV 1 to LV4 as compared to the distance from LV electrodes 40 In a situation in which two or more A-V delays have the same
(2.3) to 1. In this example, Wis /2 since the distance from LV1 minimum FI, the lowest A-V delay is selected as the optimal
to LV4 is twice as long as distance from electrode (2.3) to A-V delay.
LV1. W can be adjusted depending upon the LV medical To evaluate an optimal A-V delay changes during atrial
electrical lead and the spacing used between the plurality of pacing, atrial pacing is initiated at a rate equal to or just above
electrodes thereon. 45
The FI equation that is used to calculate the weighted the patient's resting sinus rate. Table 2 Summarizes fusion
fusion index for a given value A-V delay depends on the indices (FI) at differently paced A-V delays (PAV) that have
optimal LV electrode that is selected. For example, if LV1 is been computed uS1ng a similar method as that which 1S
the optimal LV electrode, then FI(1) is used to calculate the FI described relative to SAV. The optimal PAV in this case is 160
for each of the A-V delays that are being tested. If LV2 is the 50 ns. Since both PAVs of 160 and 180 ms have the same fusion
optimal LV electrode then FI(2) is used to calculate and index, the lesser of the two PAVs is selected.
optimize the A-V delay. If LV3 is the optimal LV electrode The AAV, is the difference between optimal PAV and
then FI(3) is used to calculate and optimize the A-V delay. If optimal SAV and is noted as follows:
LV4 is the optimal LV electrode then FI(4) is used to calculate
and optimize the A-V delay. 55 AAV, optimal PAV-optimal SAV=(160-120)ms=40
After a FI has been determined for the first A-V delay, the S.
programmer 24 automatically selects a second A-V delay.
Again, pacing pulses are delivered through the LV electrode Atrial pacing can also be initiated at decreasing cycle
at a second A-V delay while the sensing LV electrodes sense lengths in steps of 50 ms from the resting cycle-length. The
at operation 308. The activation times for the non-pacing LV 60 same procedure can be repeated in order to identify the opti
electrodes are then measured for the second A-V delay at mal PAV at each cycle-length. For example, the lowest FI is
operation 310. The FI for the second A-V delay is calculated identified and then the corresponding PA-V is selected. The
at operation 312 using the same FI equation that was used to corresponding optimal SAV for each cycle-length may be set
calculate the FI for the first A-V delay. After determining the by subtracting AAV, from the optimal PAV at that cycle
second FI for a second A-V, the programmer 24 automatically 65 length. The range of cycle-lengths covered in this manner
selects a third A-V delay and then the programmer 24 sends may start from the resting cycle-length and end in the upper
pacing pulses to the RV electrode or the LV electrode. A third atrial tracking rate.
US 8,929,984 B2
21 22
TABLE 2
FI for a range of PAV delays at atrial pacing with cycle
length equal or iust above the resting heart rate
PA-V delay (ms
40 6O 80 100 120 140 16O 18O 200 220 240 260
FI (ms) 39 39 39 32 30 28 25 25 28 34 40 46
Table 3 is a look-up table of optimal PAV and SAV values lar pacing lead or Boston Scientific’s EASYTRAK left ven
for different cycle-lengths that can be used for optimal and tricular pacing/sensing lead can be used.
dynamic adaptation of A-V delay corresponding to different The techniques described in this disclosure, including
sensed or paced cycle-lengths. In particular, A-V optimiza 15 those attributed to the IMD 16, the programmer 24, or various
tion can automatically adjust A-V delays according to constituent components, may be implemented, at least in part,
changes in heart rates (e.g. faster heart rates or shorter cycle in hardware, Software, firmware, or any combination thereof.
lengths). The programmer 24 or IMD 16 can adjust the AV For example, various aspects of the techniques may be imple
delay by using the look-up table that relates cycle length, PAV mented within one or more processors, including one or more
and/or SAV. For example, the IMD 16 can easily adjust the AV microprocessors, DSPs, ASICs, FPGAs, or any other equiva
delay (whether atrial-sensed or atrial-paced) according to the lent integrated or discrete logic circuitry, as well as any com
detected current cycle-length. Referring briefly to Table 3, binations of Such components, embodied in programmers,
cycle length 750 ms corresponds to a PAV of 180 ms and a Such as physician or patient programmers, stimulators, image
SAV of 160 ms. Accordingly, the PAV can be adjusted or the processing devices, or other devices. The term “module.”
SAV can be adjusted to the designated optimum levels. 25 “processor, or “processing circuitry may generally refer to
Table 3 is automatically generated by the programmer 24 any of the foregoing logic circuitry, alone or in combination
and stored into memory. Programmer 24, for example, can with other logic circuitry, or any other equivalent circuitry.
initiate atrial pacing at different rates. The optimal PAV can be Such hardware, software, and/or firmware may be imple
determined and stored into memory for a given atrial pacing 30
mented within the same device or within separate devices to
rate. The corresponding optimal SAV can be determined for Support the various operations and functions described in this
the same rate by subtracting AAV, as previously discussed disclosure. In addition, any of the described units, modules, or
and storing the optimal SAV value for that rate. components may be implemented together or separately as
Table 3 is a look-up table of optimal PAV and SAV for differ discrete but interoperable logic devices. Depiction of differ
ent cycle-lengths from resting (1000 ms) to upper tracking 35 ent features as modules or units is intended to highlight dif
rate (500 ms) ferent functional aspects and does not necessarily imply that
Such modules or units must be realized by separate hardware
or software components. Rather, functionality associated
Cycle length (CL) Optimum PAV Optimum SAV with one or more modules or units may be performed by
(ms) (ms) (ms) 40 separate hardware or Software components, or integrated
1OOO
950
18O
18O
140
140
within common or separate hardware or software compo
nentS.
900 160 120
8SO 160 120 Skilled artisans also appreciate that the exemplary methods
800 160 120 presented in the flow diagrams are intended to illustrate the
750
700
160
160
120
120
45 general functional operation of the devices described herein,
6SO 140 100
and should not be construed as reflective of a specific form of
600 140 100 software or hardware necessary to practice all of the methods
550 120 8O described herein. It is believed that the particular form of
500 120 8O software will be determined primarily by the particular sys
50 tem architecture employed in the device (e.g., IMD 16, pro
While the invention has been described in its presently grammer 24) and by the particular detection and therapy
preferred form, it will be understood that the invention is delivery methodologies employed by the device and/or sys
capable of modification without departing from the spirit of tem. Providing software and/or hardware to accomplish the
the invention as set forth in the appended claims. For described methods in the context of any modern IMD or
55 programmer, given the disclosure herein, is within the abili
example, in one or more embodiments, two or more LV ties of one of skill in the art.
electrodes may be selected for multi-site pacing of the LV. An When implemented in software, the functionality ascribed
example of such a configuration may be seen with respect to to the systems, devices and techniques described in this dis
U.S. Pat. No. 6,804,555 issued Oct. 12, 2004, and assigned to closure may be embodied as instructions on a computer
the assignee of the present invention, the disclosure of which 60 readable medium such as RAM, ROM, NVRAM, EEPROM,
is incorporated by reference in its entirety herein. Moreover, FLASH memory, magnetic data storage media, optical data
while the electrodes have been described as being able to storage media, or the like. The instructions may be executed
either sense or pace, skilled artisans appreciate that other by one or more processors to support one or more aspects of
embodiments can employ electrodes that are able to both the functionality described in this disclosure. It is appreciated
sense and pace. Additionally, many different medical electri 65 that the LV electrodes can be placed at locations about and/or
cal leads can be used to implement one or more embodiments. along the LV. It is also appreciated that more than four LV
For example, St Jude's QuartetTM Quadripolar, left-ventricu electrodes can be used to monitor electrical activation times.
US 8,929,984 B2
23 24
Furthermore, it is understood that FI is a function of mul 6. The method of claim 5 further comprising:
tiple variables Such as pacing electrode, A-V delay. Optimi determining a weighted fusion index for each of N number
zation of FI is based on any one variable while keeping the of A-V delays; and
other variables at a constant value. Additionally, other comparing weighted fusion indices for the NA-V delays,
embodiments are contemplated in which a physician may wherein N is any number between 2 and 100.
optionally perform one or more operations for any methods 7. The method of claim 1 further comprising:
described herein. eliminating either the first or the second A-V delays based
This disclosure has been provided with reference to illus upon whichever A-V delay is associated with a greater
trative embodiments and is not meant to be construed in a
limiting sense. As described previously, one skilled in the art 10
value of weighted fusion index.
will recognize that other various illustrative applications may 8. The method of claim 7 wherein the optimal A-V delay is
use the techniques as described herein to take advantage of a last remaining A-V delay that has not been eliminated out of
the beneficial characteristics of the apparatus and methods NA-V delays.
described herein. For example, it is contemplated that other 9. The method of claim 1 further comprising:
embodiments could use electrodes that are configured to pace 15 eliminating either the first A-V delay or the second A-V
and sense. Various modifications of the illustrative embodi delay based upon whichever A-V delay has a greater
ments, as well as additional embodiments of the disclosure, value when the weighted fusion indices for the first and
will be apparent upon reference to this description. the second A-V delays are equal.
10. The method of 9 wherein the first and second A-V delay
What is claimed: are associated with atrial sense at a resting atrial rate.
1. A method of cardiac pacing employing a right ventricu 11. The method of claim 10 wherein the resting atrial rate
lar electrode and a plurality of left ventricular electrodes, is between about 50 to about 90 beats per minute.
comprising: 12. The method of claim 9 wherein the first and second A-V
a) pacing a left ventricular electrode using a first A-V delay are associated with atrial pacing performed at a rate
delay; 25 greater than or equal to a resting atrial rate by less than 5 beats
b) measuring activation times at other ones of the left per minute and less than or equal to a programmed atrial
Ventricular electrodes and at one or more right ventricu upper tracking rate.
lar electrodes in response to pacing the left ventricular 13. The method of claim 12 further comprising determin
electrode: ing AAV, by Subtracting an optimal sensed A-V delay from
c) determining a weighted fusion index associated with the 30 an optimal paced A-V delay for the rate which is greater than
first A-V delay; or equal to the resting rate by less than 5 beats per minute.
d) pacing the left ventricular electrode using a second A-V 14. The method of claim 13 whereinthe optimal paced A-V
delay; delay is determined for each of N number of atrial pacing
e) measuring activation times at other ones of the left rates above the resting atrial rate but less than or equal to a
Ventricular electrodes and at the one or more right ven 35 programmed atrial upper tracking rate.
tricular electrodes; 15. The method of claim 14 further comprising determin
f) determining a weighted fusion index for the second A-V ing an optimal sensed A-V delay for each corresponding atrial
delay; rate.
g) comparing weighted fusion indices for the first and the 16. The method of claim 15 wherein the optimal sensed
second A-V delays; and 40 A-V delay is determined by subtracting AAV,Aest from the
h) selecting an optimal A-V delay in response to comparing optimal paced A-V delay.
weighted fusion indices for the first and the second A-V 17. The method of claim 15 wherein N ranges from 2-100.
delays, wherein the weighted fusion indices are deter 18. The method of claim 15 wherein a look-up table of
mined using: optimal paced and sensed A-V delays for different atrial rates
FI(A)=X "w,(i,j)|LVAT(i.A)|+X "w (i,j) IRVAT(i. 45 may be stored and paced and sensed A-V delays are adapted
A), wherein “n” is a total number of left ventricular accordingly in response to changes in the atrial rates during
electrodes, m is the total number of right ventricular Subsequent delivery of cardiac resynchronization therapy.
electrodes, LVAT is left ventricular activation times, 19. A system of cardiac pacing employing a right ventricu
RVAT is right ventricular activation times, “i' is a non lar electrode and a plurality of left ventricular electrodes,
pacing electrode and represents the pacing electrode. 50 comprising:
2. The method of claim 1 wherein the weighted fusion a) means for pacing a left ventricular electrode using a first
index include a weighting factor associated with the left ven A-V delay;
tricular electrode of the plurality of left ventricular electrodes b) means for measuring activation times at other ones of the
and another weighting factor associated each electrode of the left ventricular electrodes and at one or more right ven
one or more RV electrodes, each weighting factor is based on 55 tricular electrodes in response to pacing the left ven
each electrode's distance from the pacing electrode. tricular electrode:
3. The method of claim 2 wherein a summation of each c) means for determining a weighted fusion index associ
weighting factor associated with each electrode of the left ated with the first A-V delay;
Ventricular electrodes and another weighting factor for each d) means for pacing the left ventricular electrode using a
electrode of the one or more RV electrodes is 1. 60 second A-V delay;
4. The method of claim 1 further comprising: e) means for measuring activation times at other ones of the
determining that the optimal A-V delay is associated with left ventricular electrodes and at the one or more right
a lowest value of weighted fusion index. ventricular electrodes;
5. The method of claim 4 further comprising determining f) means for determining a weighted fusion index for the
that the optimal A-V delay is associated with a smallest value 65 second A-V delay;
A-V delay if the first and second A-V delays are associated g) means for comparing weighted fusion indices for the
with equal values of fusion index. first and the second A-V delays; and
US 8,929,984 B2
25 26
h) means for selecting an optimal A-V delay in response to 27. The system of claim 24 wherein a look-up table of
comparing weighted fusion indices for the first and the optimal paced and sensed A-V delays for different atrial rates
second A-V delays, wherein the weighted fusion indices may be stored and paced and sensed A-V delays are adapted
are determined using: accordingly in response to changes in the atrial rates during
FI(, A) X "w (i,j)|LVAT(i.A)+X, "w (i,j)|RVAT(i. 5 Subsequent delivery of cardiac resynchronization therapy.
A), wherein “n” is a total number of left ventricular 28. The system of claim 23 further comprising:
electrodes, m is the total number of right ventricular means for eliminating either the first or the second A-V
electrodes, LVAT is left ventricular activation times, delays based upon whichever A-V delay is associated
RVAT is right ventricular activation times, “i' is a non with a greater value of fusion index.
pacing electrode and represents the pacing electrode. 10
29. The system of claim 28 wherein the optimal A-V delay
20. The system of claim 19 wherein the weighted fusion is a last remaining A-V delay that has not been eliminated out
index include a weighting factor associated with the left ven of NA-V delays.
tricular electrode of the plurality of left ventricular electrodes 30. The system of claim 29 wherein the optimal sensed
and another weighting factor associated each electrode of the A-V delay is determined by subtracting AAV,Aest from the
one or more RV electrodes, each weighting factor is based on 15
optimal paced A-V delay.
each electrode's relative distance from the pacing electrode.
21. The system of claim 20 wherein a summation of each 31. The system of claim 23 further comprising:
weighting factor associated with each electrode of the left means for eliminating either the first A-V delay or the
Ventricular electrodes and another weighting factor for each second A-V delay based upon whichever A-V delay has
electrode of the one or more RV electrodes is 1. a greater value when the weighted fusion indices for the
22. The system of claim 20 further comprising: first and the second A-V delays are equal.
means for determining that the optimal A-V delay is asso 32. The system of claim 23 wherein the optimal paced A-V
ciated with a lowest value of weighted fusion index. delay is determined for each of N number of atrial pacing
23. The system of claim 22 further comprising: rates above the resting atrial rate but less than or equal to a
means for determining that the optimal A-V delay is asso 25 programmed atrial upper tracking rate.
ciated with a smallest value A-V delay if the first and 33. The system of claim 22 wherein the first and second
second A-V delays are associated with equal values of A-V delay are associated with atrial sense at a resting atrial
fusion index. rate.
24. The system of claim 23 further comprising: 34. The system of claim 33 wherein the resting atrial rate is
means for determining a weighted fusion index for N num 30
between about 50 to about 90 beats per minute.
ber of A-V delays; and 35. The system of claim 33 wherein the first and second
means for comparing weighted fusion indices for the N A-V delay are associated with atrial pacing performed at a
A-V delays, rate greater than or equal to a resting atrial rate and less than
wherein N is any number between 2 and 100. or equal to a programmed atrial upper tracking rate.
25. The system of claim 24 further comprising means for 35
36. The system of claim35 further comprising determining
determining an optimal sensed A-V delay for each corre AAV, by Subtracting an optimal sensed A-V delay from an
sponding atrial rate. optimal paced A-V delay.
26. The system of claim 24 wherein N number ranges from
2 to about 100. k k k k k

US8929984

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

US8929984

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

US8929984

Uploaded by

Copyright:

Available Formats

USOO8929.

(12) United States Patent (10) Patent No.: US 8,929,984 B2

Pace at same SAWFPAW

Measure activation times atother

(56) References Cited 2013/0218225 A1 8, 2013 Ghosh et al. .................... 607/28

Measure activation times at 202 200

Determine Ffor One LV 208

Measure activation times at 202 200

Determine F for One LV 208

Difference Eliminate electrode

RVAT(1A) 400 mS WindoW

atrial event Ventricular pace (at LV1)

You might also like

US8929984

Uploaded by

Document Informationclick to expand document information........................................................................................

Document Informationclick to expand document information

Copyright:

Available Formats

US8929984

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

US8929984

Uploaded by

Copyright:

Available Formats

USOO8929.

(12) United States Patent (10) Patent No.: US 8,929,984 B2

Pace at same SAWFPAW

Measure activation times atother

(56) References Cited 2013/0218225 A1 8, 2013 Ghosh et al. .................... 607/28

Measure activation times at 202 200

Determine Ffor One LV 208

Measure activation times at 202 200

Determine F for One LV 208

Difference Eliminate electrode

RVAT(1A) 400 mS WindoW

atrial event Ventricular pace (at LV1)

You might also like