(10) Patent No.:
`a2, United States Patent
`US 6,527,729 BI
`(45) Date of Patent:
`Turcott
`Mar.4, 2003
`
`
`US006527729B1
`
`(54) METHOD FOR MONITORING PATIENT
`USING ACOUSTIC SENSOR
`
`(75)
`
`Inventor: Robert Turcott, Mountain View, CA
`(US)
`
`.
`(73) Assignee: Pacesetter, Inc., Sunnyvale, CA (US)
`(*) Notice:
`Subject to any disclaimer, the term ofthis
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 129 days.
`
`4,446,873 A *
`4,763,646 A
`4,905,706 A *
`4,989,611 A *
`5,042,497 A
`5,168,869 A
`5,404,877 A
`5,554,177 A
`5,862,805 A
`5,935,081 A
`
`5/1984 Groch et al. wee 600/528
`8/1988 Lekholm ............ 128/419 PG
`3/1990 Duff et al. we. 600/514
`2/1991 Zanetti et al. oc. 600/508
`
`8/1991 Shapland .......eeeeeenes 128/696
`« 128/419 PG
`12/1992 Chirife ........
`
`s.scssseessseeee 128/671
`4/1995. Nolan et al.
`9/1996 Kieval et al. sesesessee 607/17
`
`1/1999 Nitzan .......ccccececceceeee ee 128/898
`8/1999 Kadhiresan bee eeeseeseeeees 600/513
`
`(21) Appl. No.: 09/689,270
`
`(22)
`
`Filed:
`
`Oct. 11, 2000
`
`* cited by examiner
`
`oo
`Related U.S. Application Data
`(63) Continuation-in-part of application No. 09/543,394,filed on
`ABSTRACT
`67)
`Apr. 5, 2000, which is a continuation-in-part of application
`.
`.
`oo,
`No. 09/467,298,filed on Dec. 17, 1999, which is a continu-
`A method for monitoring the progression of the disease of a
`ation-in-part of application No. 09/438,017, filed on Nov.
`heart failure patient is provided. An implantable or other
`10, 1999, now Pat. No. 6,409,675.
`
`(SV) Tint. C07 eee eeeseeceeeeeeeeeeeeeeeeeneees A61B 5/04=ambulatory monitor senses acoustic signals including heart
`(52) US. C1. cece cescnseeeecnesenseecenseesenecassenees 600/528
`and lung sounds within the patient. Significant changes in
`
`
`(58) Field of Search....
`.. 600/508, 514,
`the energy content of either the heart or lung sounds is
`600/528
`indicative of a heart failure exacerbation. This information
`
`Primary Examiner—Carl Layno
`(74) Attorney, Agent, or Firm—Steven M. Mitchell
`
`(56)
`
`.
`References Cited
`U.S. PATENT DOCUMENTS
`
`may be used to warn the patient or healthcare providers of
`changes in the patient’s condition warranting attention.
`
`3,985,121 A
`
`10/1976 Hellenbrand ............0.. 128/2 K
`
`42 Claims, 21 Drawing Sheets
`
`81
`
`83
`
`
`
`80
`
`82
`
`83
`
` y
`
`Sample phonocardiogram
`
`
`Store 10 sec record to
`long-term memory
`
`
` y
`
`
`Retrieve QRSlocations
`from memory
`———
`
`Sum amplitudes over
`
`
`
`Calculate energy of S1-S4
`lung sounds window
`86
`Average overall heart beats
`Normalize by
`
`
`in the present data set
`numberof samples
`85
`Store energy in
`88
`Store in
`
`
`long-term memory
`long-term memory
`
`90
`
`
`
`recent lung sound averages
`
`Calculate average and
`standard deviation of
`
`recent S,-S4
`
`
`
`
`n=#of energiesthatdiffer
`present measure
`
`
`from recent averages by
`greater than recent average
`
`
`> 1.5 x std. dev.
`
`by more than 1.5x
`std. dev?
`
`
`byn
`Incrementalert counter
`
` Increment alert counter
`
`
`
`
`
`
`Calculate average and
`standard deviation of
`
`87
`
`APPLE 1006
`
`APPLE 1006
`
`1
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 1 of 21
`
`US 6,527,729 B1
`
`/JoyUUsueL
`
`IDAIIOIY
`
`UOINNITA
`
`MMII
`
`[‘SI
`
`vi
`
`juaned
`
`WolV
`
`BO!
`
`2
`
`
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 2 of 21
`
`US 6,527,729 B1
`
`100
`
`104
`
`20L
`
`22
`
`22
`
`~*6
`Fig.2a
`36>22~*
`Fig. 2b
`
`}
`
`
`
`3
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 3 of 21
`
`US 6,527,729 B1
`
`22
`
`
`
`4
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 4 of 21
`
`US 6,527,729 B1
`
`40
`
`TOLDDLLLLLE.
`
`Fig. 4a
`
`‘)
`
`38
`
`x
`
`5
`
`

`

`U.S. Patent
`
`US 6,527,729 B1
`
`Mar.4, 2003
`
`Sheet 5 of 21
`
`40
`
`6
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 6 of 21
`
`US 6,527,729 B1
`
`Se
`
`12
`
`36
`
`
`
`
`
`
`( = —Y
`
`
`51 teuel\53
`
`
`
` 7ABAa
`
`7
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 7 of 21
`
`US 6,527,729 B1
`
`\\\
`
`f Fig.6
`
`120
`
`8
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 8 of 21
`
`US 6,527,729 B1
`
`OY
`
`Fig. 7a
`7b
`Fig. 7b
`
`°°
`
`20
`
`100
`
`110
`Fig. 7c
`100
`|) 36
`
`
`
`08 ‘i 108
`All\) 102
`
`CH
`
`
`
`
`
`
`
`20
`
`102. 1"
`
`.
`
`HL) 103
`
`
`
`9
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 9 of 21
`
`US 6,527,729 B1
`
`Sample ECG
`
`Store 10 sec record to]
`long term memory
`
`Identify location of each
`QRS complex
`
`60
`
`62
`
`64
`
`Store location of each
`QRS complex in memory
`
`66
`
`Calculate RR Intervals
`
`Calculate HRV measure
`
`Store HRV measure
`in long-term memory
`
`Calculate average &
`Standard deviation of
`recent HRV measures
`
`
`
`
`n= the number of HRV
`measures that differ
`from recent averages
`by morethan 1.5 x
`std. deviations
`
`
`
`
`68
`
`70
`
`72
`
`74
`
`76
`
`78
`
`Incrementalert
`counter by n
`
`Fig. 8
`
`10
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 10 of 21
`
`US 6,527,729 B1
`
`T
`
`P
`
`Sy
`
`ECG
`400
`
`Normal
`406
`
`Amplitude
`
`S3
`
` Acute
`
`CHF exacerbation
`
`Fig. 9
`
`11
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 11 of 21
`
`US 6,527,729 Bl
`
`Sample phonocardiogram
`
`Store 10 sec record to
`long-term memory
`
`Retrieve QRSlocations
`from memory
`
`Calculate energy of S;-S4
`
`80
`
`82
`
`83
`
`84
`
`86
`
`Fig. 10
`
`Sum amplitudes over
`lung sounds window
`
`Average overall heart beats
`in the present data set
`
`Normalize by
`number of samples
`
`Store energy in
`long-term memory
`
`8
`
`8
`
`Store in
`long-term memory
`
`81
`
`83
`
`5
`
`Calculate average and
`standard deviation of
`
`recent S,-S4
`
`90
`
`Calculate average and
`standard deviation of
`recent lung sound averages
`
`87
`
`
`92
`
`
`
`n = # of energiesthat differ
`present measuré
`
`from recent averages by
`greater than recent average
`
`
`
`> 1.5 x std. dev.
`by more than 1.5x
`
`
`std. dev?
`93-7 End)
`p74 4
`Increment alert counter
`by n
`
`Incrementalert counter
`
`12
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 12 of 21
`
`US 6,527,729 B1
`
`=—
`
`— O
`
`D
`eg
`
`©E
`
`S
`wT
`
`NS
`—_
`
`a
`
`©—
`
`ECG
`
`amplitude
`
`13
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 13 of 21
`
`US 6,527,729 Bl
`
`of signal in window
`following the
`QRS complex
`
`Sample
`plethysmographysignal
`162
`164
`For each QRSpreviously
`identified and stored
`in memory,
`do the following:
`
`
`
`
`
` Obtain maximum value
`
`
`
`
` Obtain minimum value
`
`of the signal in the
`
`
`window following
`the QRS complex
`
`
`Calculate
`
`the pulse amplitude
`
`170
`
`
`174
`
`
`QRS complexes
`been processed
`
`
`?
`
`
`Fig. 12 ™
`
`176
`
`180
`
`18
`
`Store first 60 sec to
`long-term memory
`178
`
`Calculate
`ave & std. dev
`of pulse amplitude
`
`Store to
`long-term memory
`
`Calculate
`ave & std. dev.
`of recent measures
`
`
`
`184
`
` Is current.
`
`ave pulse amplitudé
`
`
`< recent ave
`-1.5 x.std. dev
`
`of ave
`
`
`
`Increment
`alert counter
`
`
`
`pulse std. dev.
`< recent ave
`
`std. dev. -1.5 x std de
`
`of std. dev.
`
`
`
`14
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 14 of 21
`
`US 6,527,729 B1
`
`Sample
`plethysmography
`signal
`
`130
`Examine next
`data point
`
`134
`
`end
`of data
`
`
`
`136
`
`threshold
`
`
`
`
`
`Get maximum value
`in following
`
`time window
`142
`
`146
`
`Store first 60 sec to
`long-term memory
`148
`
`Calculate
`ave. & std. dev.
`of pulse amplitude
`
`150
`
`215
`
`Store ave. & std. dev.
`to long-term memory
`
`Calculate
`ave & std. dev.
`of recent measures
`
`154
`
` Is current.
`ave pulse amplitude
`< recent ave
`-1.5 x.std. dev
`of ave
`
`Increment
`alert counter
`
`
`
`
`158
`
`Is current
`
`pulse std. dev
`Y
`< recent ave
`
`Retain
`
`pulse amplitude][Tncrement
`alert counter
`Ny
`In memory
`160
`
`Get minimum value
`in preceeding
`time window
`
`pulse amplitude
`144
`
`
`
`15
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 15 of 21
`
`US 6,527,729 B1
`
`|
`apneic
`| hyperventilation|
`A10
`!
`+— phase —>
`|Nv]-"-|
`
`|
`
`lung volume
`|
`! O, saturation
`|
`
`|
`|
`|
`
`412
`
`414
`
`| |
`
`:
`
`|
`
`| | |
`
`CO,
`concentration
`
`|
`
`|
`!
`
`|
`
`amplitude
`
`!
`
`|
`
`| | t
`
`ime
`
`16
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 16 of 21
`
`US 6,527,729 B1
`
`Calculate
`maximum difference
`of MAXAG
`228
`226
`1<S OMAXAC;?
`
`
`
`N
`
`232
`Caleul
`maximumdifference
`of MAXFREQ;
`
`Sample ECG
`902
`200
`Identify location of
`each QRS complex
`
`203
`Store location of QRS
`complexes in memory
`
`Calculate RR intervals
`
`204
`
`224
`
`210
`Set pointer to
`first interval
`pointer
`208
`
`Fill array with N intervals,
`starting with pointer
`
`Calculate FFT of array
`
`
`
`> OMAXFREQ?
`236
`AVERR;= dc value of FFT Increment|IN
`
`
`
`234
`
`
`
`238
`
`Calculate
`maximum difference
`of AVERR;
`
`240
`
`?
`
`%<
`
`> OAVERR
`242
`Increment
`alert counter
`
`N
`
`24, End)
`Fig. 15
`
`17
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 17 of 21
`
`US 6,527,729 B1
`
`Load array
`with pulse amplitudes
`
`266
`
`Append @sas
`necessary
`to fill array
`
`270
`Calculate FFT of array
`
`272\
`
`TR
`Find maximum
`ac value
`
`274
`
`corresponding
`frequency
`>0.005 and <0.02
`
`
`
`
` 278
` Is
`
`max ac value
`7 Up!
`
`282
`CEnd)
`
`9
`
` Y
`
`Increment
`
`280
`
`Fig. 16
`
`Sample
`plethysmography
`signal
`
`252
`250
`For each QRS
`previously identified
`and stored in memory,
`do the following:
`
`Obtain the maximum
`value that the signal
`attains ina
`predetermined window
`following the QRS
`complex
`
`
`
`
`
`Obtain the minimum
`value that the signal
`attains in a
`predetermined window
`following the QRS
`complex
`
`262
`
`Calculate
`pulse amplitude
`264
`
`Retain
`pulse amplitude
`in memory
`
`
`
`18
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 18 of 21
`
`US 6,527,729 B1
`
`Sample the
`O, signal
`
`Calculate
`running average
`
`Find maximum
`O, value
`
`Find minimum
`O, value
`
`300
`
`302
`
`304
`
`306
`
`Obtain difference
`
`308
`
`3 10
`
`difference > 09)?
`
`>
`
`
`Y
`
`312
`
`Increment
`alert counter
`
`314 End)
`
`Fig. 17
`
`19
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 19 of 21
`
`US 6,527,729 B1
`
`QTSI
`
`
`
`OCHopnitdure
`out} OPTttreastnd
`VoVuonemmes°OwTOOOO—AADY
`DOoye1yIRay
`
`
`20
`
`20
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 20 of 21
`
`US 6,527,729 B1
`
`21
`
`21
`
`

`

`U.S. Patent
`
`Mar.4, 2003
`
`Sheet 21 of 21
`
`US 6,527,729 B1
`
`Sample O, sensor
`
`462
`
`Average samples
`
`464
`
`466
`
`Is average < threshold?
`
`
`
`7 470 End}
`
`468
`
`Increment
`alert counter
`
`Fig. 20
`
`22
`
`

`

`US 6,527,729 B1
`
`1
`METHOD FOR MONITORING PATIENT
`USING ACOUSTIC SENSOR
`
`This is a continuation-in-part of application Ser. No.
`09/543,394,filed Apr. 5, 2000, which is a continuation-in-
`part of application Ser. No. 09/467,298,filed Dec. 17, 1999,
`which is a continuation-in-part of application Ser. No.
`09/438,017, filed Nov. 10, 1999, now US. Pat. No. 6,409,
`675.
`
`BACKGROUND OF THE INVENTION
`
`I. Field of the Invention
`
`This invention relates generally to implantable monitor-
`ing devices, and more particularly to a method for monitor-
`ing the status of a patient- with a chronic disease such as
`heart failure using heart and lung sounds.
`II. Description of the Related Art
`Manychronic diseases, such as diabetes and heartfailure,
`require close medical management to reduce morbidity and
`mortality. Because the disease status evolves with time,
`frequent physician follow-up examinations are often neces-
`sary. At follow-up, the physician may make adjustments to
`the drug regimenin order to optimize therapy. This conven-
`tional approach of periodic follow-up is unsatisfactory for
`some diseases, such as heart failure, in which acute, life-
`threatening exacerbations can develop between physician
`follow-up examinations. It is well know among clinicians
`that if a developing exacerbation is recognized early, it can
`be easily and inexpensively terminated,
`typically with a
`modest increase in oral diuretic. However, if it develops
`beyond the initial phase, an acute heart failure exacerbation
`becomesdifficult to control and terminate. Hospitalization in
`an intensive care unit is often required.It is during an acute
`exacerbation of heart failure that many patients succumb to
`the disease.
`
`It is often difficult for patients to subjectively recognize a
`developing exacerbation, despite the presence of numerous
`physical signs that would allow a physician to readily detect
`it. This problem is well illustrated by G. Guyatt in his article
`entitled “A 75-Year-Old Man with Congestive Heart
`Failure,” 1999, JAMA 281(24)2321-2328. Furthermore,
`since exacerbations typically develop over hours to days,
`even frequently scheduled routine follow-up with a physi-
`cian cannot effectively detect most developing exacerba-
`tions. It is therefore desirable to have a,.system that allows
`the routine, frequent monitoring of patients so that an
`exacerbation can be recognized early in its course. With the
`patient and/or physician thus notified by the monitoring
`system of the need for medical intervention, a developing
`exacerbation can easily and inexpensively be terminated
`early in its course.
`The multiplicity of feedback mechanismsthat influence
`cardiac performance places the heart at
`the center of a
`complex control network. The neurohumoral axis includes
`the autonomic nervous system, consisting of sympathetic
`and parasympathetic branches, and numerous circulating
`hormones such as catacholamines, angiotensin, and aldos-
`terone. Neural reflex arcs originating from pressure and
`stretch receptors, which directly measure mechanical hemo-
`dynamic status, modulate the neurohumoral axis. Similarly,
`chemoreceptors respond to changes in CO2, pH, and O2,
`which assesses cardiopulmonary function. The neurohu-
`moral system influences cardiac performanceat the level of
`the cardiac electrical system by regulating heart rate and the
`conduction velocity of electrical depolarizations.
`It also
`influences cardiac performance at the mechanicallevel, by
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`the effective vigor with
`that is,
`controlling contractility,
`which the heart muscle contracts. Conventional cardiac
`monitors, such as defibrillators, pacemakers, Holter
`monitors, and cardiac event records, are tailored for the
`diagnosis and/or therapy of abnormalities of the cardiac
`electrical system. In contrast, heart failure is a disease of the
`cardiac mechanical system: it is primarily a failure of the
`myocardium to meet
`the mechanical pumping demands
`required of it. In monitoring the status of a heart failure
`patient, measuring the mechanical hemodynamic variables
`is clearly desirable. Examples of mechanical hemodynamic
`variables include atrial, ventricular, and arterial pressures,
`and cardiac output (volume of blood pumpedinto the aorta
`per unit time). However, because of the complex feedback
`network that monitors and controls cardiac performance,
`measuring variables that do not directly reflect the mechani-
`cal performance of the heart is also useful. In this way,
`measuring cardiac electrical activity to assess heart rate
`variability (described below) allows one to infer the state of
`the autonomic nervous system, which allows one to infer
`information about the hemodynamicstatus of a heart failure
`patient. Similarly, recognition of Cheyne-Stokes respiration
`(described below)via respiratory pattern analysis, hemoglo-
`bin saturation analysis, and blood gas analysis allows one to
`detect the presence of pulmonary edema,and thereby detect
`an acute heart failure exacerbation, though none of these
`parameters directly measure mechanical hemodynamicsta-
`tus.
`
`One approach to frequent monitoring of heart failure
`patients that has been proposedis the daily acquisition of the
`patient’s weight and responsesto questions about subjective
`condition (Alere DayLink Monitor, Alere Medical, Inc., San
`Francisco, Calif.). The simplicity and noninvasive embodi-
`mentof this approach are desirable features. However, both
`the amount and the sophistication of the objective physi-
`ological data that can be acquired in this way are quite
`limited, which consequently limits the accuracy of the
`system. Furthermore, the system requires the active partici-
`pation of the patient, who must not deviate from the precise
`data acquisition routine or risk introducing confounding
`factors into the acquired data.
`Some of these limitations have been addressed by the
`developmentof an implantable system that monitors hemo-
`dynamic status (Medtronic Chronicle, Medtronic,
`Inc.,
`Minneapolis, Minn.). While this system potentially avoids
`the need for active patient participation,
`it relies on an
`intravascular sensorplaced in the right ventricle of the heart.
`This approachis consistent with the prior art for implantable
`hemodynamic status monitoring, which has to date focused
`on intravascular or
`intramyocardial
`instrumentation.
`Examples include US. Pat. No. 5,454,838 in which Vallana
`et al. teach placement of a sensor on the myocardial wall
`using an intravascular approach. In U.S. Pat. No. 5,496,351,
`Plicchiet al. propose placing a sensor within the myocardial
`wall. Mortazavi in U.S. Pat. No. 5,040,538 and Cohenetal.
`in U.S. Pat. No. 4,815,469 describe placementof an optical
`sensor within the right ventricle. In the context of hemody-
`namic assessment for arrhythmia discrimination, Cohen and
`Liem (Cire., 1990, 82:394-406) study the effectiveness of a
`pressure transducer placed in the right ventricle. Clearly,
`powerful
`information about hemodynamic status can be
`obtained using intravascular instrumentation. However,
`intravascular or intramyocardial instrumentation carries sig-
`nificant risks to the patient, including increased periopera-
`tive morbidity and mortality, and increased long-term risks
`such as stroke and pulmonary embolism. Furthermore, intra-
`vascular instrumentation can only be performed by exten-
`23
`
`23
`
`

`

`US 6,527,729 B1
`
`3
`sively trained specialists, thereby limiting the availability of
`qualified physicians capable of implanting the device, and
`increasing the cost of the procedure. Finally, because of the
`added patient risks and greater physical demands of an
`intravascular environment, the intravascular placement of
`the sensor increasesthe cost of development, manufacturing,
`clinical trials, and regulatory approval.
`Though not directly related to hemodynamic status
`monitoring, extravascular sensing of cardiac electricalactiv-
`ity is known in the art. Early generations of implantable
`pacemakersand defibrillators relied on epicardial placement
`of sensing electrodes. Epicardial electrodes still see use in
`special patient populations. Extrathoracic sensing of cardiac
`electrical activity is also possible, which avoids the need for
`direct contact with the heart, and thus decreasesthe difficulty
`of the implant procedure and reduces the risk of periopera-
`tive complications. An example of this approach is the
`Reveal
`Insertable Loop Recorder
`(Medtronic,
`Inc.,
`Minneapolis, Minn.), a cardiac event recorder configured for
`short-term implantation. As a temporarily implantable
`recorder,
`it overcomes some of the technical difficulties
`associated with conventional externally worn recorders of
`cardiac electrical activity. Two general types of externally
`worn recorders are Holter monitor recorders, which record
`continuously for an extended period of time, and cardiac
`event recorders, such as the King of Hearts (Alaris Medical
`Systems, San Diego, Calif.), which use a loop memory to
`retain the most recent history of cardiac electrical activity.
`Both these approaches require surface contact electrodes
`which are cumbersomeand inconvenient for the patient, and
`more susceptible to motion artifact than an implanted elec-
`trode. However, like conventional cardiac event recorders
`and continuous Holter monitor recorders, the Reveal Insert-
`able Loop Recorder is designed for short-term use as a
`diagnostic aid. More importantly, it requires active patient
`participation; when the patient recovers from a syncope, or
`becomes aware of symptoms, he must signal to the event
`recorder by meansof an Activatorthat the recent data should
`be retained in long-term memory for later review by a
`physician. After diagnosis the Reveal
`Insertable Loop
`Recorder is explanted from the patient. Thus the Reveal is
`intended for short-term recording for diagnostic use,
`is
`limited to recording the electrical activity of the heart, and
`does not attempt to measure or quantify the hemodynamic
`status of the patient beyond screening for cardiac arrhyth-
`mias.
`
`An extension of the short-term recorders just described is
`the Implantable Ambulatory Electrocardiogram Monitor
`described by Nappholz et al. in U.S. Pat. No. 5,113,869,
`incorporated herein by reference. This device is designed for
`chronic extravascular implantation. In contrast to cardiac
`recorders,
`it performs analysis on the electrocardiogram
`signal in order to predict imminent cardiac arrhythmias and
`to detect cardiac ischemia. Like the cardiac recorders,it is
`capable of storing raw ECG data for later review by a
`physician. This feature, along with the record of arrhythmic
`events it detected, allows the physician to tailor pharmaco-
`logic therapy. In addition, Nappholz et al. mention the use of
`transthoracic impedance for minute ventilation, ultrasound
`transducers for arterial pressure, or other sensors to allow
`discrimination of arrhythmias from normal cardiac rhythms
`caused by exertion or physiologic stress.
`While the Holter monitor recorder, the Reveal Insertable
`Loop Recorder, and the Implantable Ambulatory Electro-
`cardiogram Monitor provide important clinical utility in
`recording and monitoring cardiac electrical activity, none is
`designed to monitor hemodynamic status. Indeed, cardiac
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4
`electrical activity does not, by itself, provide unambiguous
`information about hemodynamic status. By sensing only
`cardiac electrical activity, these devices are unable to dis-
`tinguish between, for example, a hemodynamically stable
`cardiac rhythm and Pulseless Electrical Activity (PEA), a
`condition in which the heart is depolarizing normally, and
`thus generating a normalelectrical pattern, but is not pump-
`ing blood. Furthermore, these devices are unable to recog-
`nize or quantify subtle changes in the patient’s hemody-
`namic status. What
`is needed is an extravascular,
`hemodynamic monitor designed for chronic use.
`While muchofthe prior art has focused on intravascular
`instrumentation, as discussed above, some proposal has been
`made to incorporate physiologic sensorsinto the implantable
`cardiac device itself. Fearnot in U.S. Pat. No. 5,040,533
`teaches placementof a generalized window in the housing of
`the cardiac device. The window might be transparent to
`facilitate the transmission of light or flexible to facilitate
`pressure transduction. While the convenience, from the
`clinician’s perspective, of incorporating the sensors into the
`housing of the cardiac device is an obvious advantage, the
`technical difficulty in maintaining a hermetic seal between
`two different materials, particularly in a chronically
`implanted device, is equally obviousto one skilled in the art.
`The technical challenge is made moredifficult by the greatly
`increased circumference, relative to that of standard feed-
`through connections known in the art, of the boundary
`between the window and the device housing. What
`is
`needed, therefore, is a method of placing a hemodynarnic
`sensor in or on the device without compromising the integ-
`rity of the hermetic enclosure.
`Prutchi et al., in U.S. Pat. No. 5,556,421 propose place-
`mentof a sensor within the headerof a cardiac device. While
`this is an obvious solution for devices that have external
`leads requiring headers, it presupposes the existence of a
`header, and therefore does not address the implantable
`device that lacks a header. Furthermore, while appending a
`header to one end or pole of an implantable device is an
`efficient solution when external leads are required, append-
`ing a header-like sensor unit to one end or pole of a device
`not otherwise requiring a header, where the sensor unit is
`itself, like a header, the full-thickness of the device, is an
`inefficient use of volume. Thus, the approach of Prutchietal.
`used in a device that doesn’t otherwise require a header
`would be to append a header or a header-like sensor unit to
`one end or pole of the device, but this would unnecessarily
`increase both the volume and the expense of the device. A
`further disadvantage of placing a sensor in a headeris that
`it does not necessarily provide for the optimal signal trans-
`duction of a particular sensor. For example, the performance
`of the optical sensor described in the above referenced U.S.
`Pat. No. 5,556,421 would be so severely degraded by direct
`transmission of light from source to detector that one skilled
`in the art would question the functionality of the proposed
`solution. In addition, placement in a rigid epoxy headeris
`simply not an option for some sensors, such as sound
`sensors, because of the dramatic degradation in the signal-
`to-noiseratio the rigid header would impose. What is needed
`is a method of incorporating a hemodynamic sensor into a
`implantable device, providing it optimal accessto the exter-
`nal milieu so that
`the signal of interest
`is optimally
`transduced, maintaining the hermetic enclosure provided by
`the device housing, and minimizing the added volumethat
`the sensor imposes.
`A solution to this challenge is offered in U.S. Pat. No.
`5,404,877 by Nolanet al., in which an arrhythmia detection
`and warning system is described. The monitor avoids exter-
`24
`
`24
`
`

`

`US 6,527,729 B1
`
`5
`leads and sensors by generating a radio frequency
`nal
`electromagnetic field within the device, whichis intended to
`propagate through the device housing, reflect off internal
`organs and structures, and be detected again inside the
`device housing. The device uses observed changesin imped-
`ance seen at the antenna to deduce organ motion, in par-
`ticular heart motion. While the leadless embodiment
`is
`
`the
`desirable for the reasons described by Nolan et al.,
`technical challenges associated with inferring useful physi-
`ologic information from changes in impedanceis obviousto
`one skilled in the art. The general problem is that the large
`number of confounding factors, e.g., changes in body
`position, would certainly swamp the subtle impedance
`changes that might result from changes in cardiac volume
`with contraction.
`
`Another aspect of the prior art that has been limited is in
`the communication of information between a device and the
`clinician. During periodic follow-up in the physician’s
`office, conventional implanted devices such as pacemakers
`and defibrillators are electronically interrogated and stored
`data is conveyed outside the body using telemetry. Depend-
`ing on the physician’s assessment, programmable device
`parameters or the patient’s medical regimen may be modi-
`fied. The processis initiated by the clinician and requires the
`placement of an external telemetry antenna in close prox-
`imity to the implanted device. Indeed,
`in U.S. Pat. No.
`5,342,408 DeCoriolis et al. provide a telemetry signal
`strength indicator that facilitates the positioning of the
`external antenna by the clinician. While the prior art is
`sufficient for conventional cardiac devices, which typically
`only require telemetry during relatively infrequent follow-
`up visits, in cases where frequent telemetry is required it is
`desirable to have a system that does not rely on active human
`participation. With the Alere system described above, data is
`conveyed daily over telephone channels to a central location
`for review, a process that is initiated by the patient and
`requires interaction of the patient with the device. While in
`this case the clinician is not actively involved in the telem-
`etry process, the patient is. This approach therefore also
`precludes a fully automated system. What is needed is a
`system that provides telemetry at a distance, so that data can
`be transferred remotely without the active participation of a
`clinician or cooperation of the patient. With telemetry at a
`distance, data could be automatically transferred for review
`and analysis by a clinician or central monitor, and program-
`ming parameters of the device could be modified remotely.
`By not relying on the routine participation of patient or
`physician, such a system would be more convenient and
`reliable, would avoid the inconvenience and expense of
`in-person follow up, and would allow frequent monitoring
`and tailoring of device parameters and medical therapy as
`the patient’s disease status changes.
`In US. Pat. No. 5,113,869 to Nappholz et al., telemetry is
`provided to warn the patient or physician of an impending
`arrhythmia event. U.S. Pat. No. 5,544,661 to Davis etal.
`discloses an ambulatory patient monitor that is worn by the
`patient and provides arrhythmia analysis and wireless two-
`way data and voice communication to a central station. In
`the event of an arrhythmia, a clinician can actively and
`remotely monitor the patient. However,
`this system is
`intended for short term use to monitor patients who may be
`subject to sudden life threatening arrhythmic events. Nap-
`pholz et al. in US. Pat. No. 5,720,770 discloses a system
`including an implanted cardiac therapy device and an exter-
`nal portable device in constant or periodic telemetric com-
`munication with the implanted device. The external device
`receives updates of the condition of the patient and the
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`6
`operation of the cardiac therapy device. This information is
`transmitted by the external device over wireless or wired
`phonelines to a central monitoring station.
`Oneof the challenges of providing telemetry at a distance
`is to provide for the efficient transmission and reception of
`energy by the implanted device. The current art places the
`telemetry coil inside the implantable cardiac device housing
`and uses magnetic inductive coupling to convey data
`between the implanted and external units. The metallic
`housing attenuates the magneticfield, but since the clinician
`is available to actively position the external coil the degree
`of attenuation is tolerable. The above referenced U'S. Pat.
`
`No. 5,404,877 describes radio-frequency electromagnetic
`telemetry, rather than the conventional magnetic-induction
`method commonly used in pacemakers and implantable
`defibrillators. However,like the conventional magneticcoil,
`the RF antenna is placed within the device housing, which
`has the undesirable effect of attenuating the signal strength.
`The above referenced U.S. Pat. No. 5,113,869 discloses a
`radio frequency telemetry system similarto that described in
`above referenced U.S. Pat. No. 5,404,877, but with the
`antenna placed outside the device housing on a lead that
`extends away from the device. The configurationis desirable
`in that attenuation by the metallic housing of the device is
`avoided, however,
`it requires subcutaneous tunneling for
`placement, which causes tissue trauma and increases the
`risk, both acute and chronic, of infection. Furthermore,
`patient motion will alter the impedance between the antenna
`and ground plane, which degrades antenna gain. The above
`referenced U.S. Pat. No. 5,342,408 teaches placement of an
`antenna in the device header, which has the advantage of
`avoiding the attenuation of the metallic housing, as well as
`avoiding the disadvantages of an antenna that extends away
`from the device in a lead. However, placement in the header
`presupposes the existence of external
`leads requiring a
`header, which are not necessarily present in a device that
`uses extravascular sensors.
`It
`is desirable,
`therefore,
`to
`provide placement of the telemetry antenna outside the
`housing of a device which lacks a header.
`the antenna is
`The placement should be such that
`mechanically stabilized and electrically insulated from the
`device housing and the surroundingtissue.
`Because of the considerations described above, the prin-
`cipal object of the present invention is to provide a method
`for use of a device that monitors a patient’s hemodynamic
`status.
`
`Another object of the invention is to monitor the status of
`a chronic disease in order to optimize medical therapy.
`A further object is to monitor the status of a chronic
`disease in order to recognize and facilitate the early termi-
`nation of a developing exacerbation.
`Further objects and advantages will become apparent
`from a consideration of the ensuing description and draw-
`ings.
`
`BRIEF SUMMARY OF THE INVENTION
`
`The preferred embodiment of the present invention pro-
`vides a method for monitoring the progression of the disease
`state of a heart failure patient. A physiologic acoustic signal
`is sensed inside a patient’s body at a first time period. A
`value is calculated corresponding to the energy content of a
`portion of the acoustic signal for the first time period. This
`information is stored in memory. The acoustic signal is
`sensed at a second later time period and a value correspond-
`ing to the energy content ofthe portion of the acoustic signal
`is calculated for the second time period. Then,the calculated
`25
`
`25
`
`

`

`US 6,527,729 B1
`
`7
`value of the energy content of the acoustic signal for the
`second time period is compared with the calculated value of
`the energy content of the acoustic signal for the first time
`period and an output is provided as a function of the results
`of the comparison. Significant excursions from the prior
`baseline energy content are indicative of a heart failure
`exacerbation. This information may be used to warn the
`patient or healthcare provider of changes in the patient’s
`condition warranting attention.
`In one embodiment
`the
`energy of heart sounds is monitored and the portion of the
`acoustic signal is correlated with the QRS complex of an
`electrocardiogram