CA2190159A1 - Rate responsive cardiac pacemaker for discriminating stair climbing from other activities - Google Patents

Abstract

A method of and apparatus for pacing a patient's heart at a pacing rate dependent on patient activity and posture particularly during stair climbing. A dual chamber, rate responsive pacemaker for pacing a patient's heart includes at least one DC accelerometer mounted in the pacemaker pulse generator for implantation such that the sensitive axis of the DC accelerometer is sensitive to the effects of gravity during forward lean of the patient characteristic of stair climbing posture. The DC and AC signal outputs of the accelerometer are processed to develop a tilt signal and an activity signal. A target rate control signal is derived from the activity signal dependent on the level of activity. A stair climbing rate is selected for controlling the physiologic pacing rate between a lower and an upper pacing rate in the presence of an activity signal indicative of a patient walking rate and a tilt signal value falling within a tilt window. The target rate control signal is used to control the pacing rate if the activity signal is indicative of faster patient movement, e.g. running, or if the tilt signal is outside the tilt window indicating that the patient is either upright or prone.

Description

W096130079 (~ 01 59 ~ -A7~?4 RATE RESPONSIVE CARDIAC PA-'F.l\~AKF.R FOR DISCRIMINATING
STAIR CLIMBING FROM OTHER ACTIVIT~ES
RF.FERF.~C~F. TO RF.T ~TF.n APPr.T(~TION
Reference is made to commonly assigned co-pending U.S. patent application filed even date herewith entitled MEDICAL DEVICE EMPLOYING MULTIPLE DC
AC(~F.T .F.Rr)MFTERS FOR PATIENT ACTIVITY AND POSTURE SENSING filed on even date herewith).
l~A(`TC('IROUNn OF T~F. INVF~TION
FiPl~l of the TnVrAnti~n The present invention relates to rate responsive cardiac I ' and more particularly to the use of a DC ac.,~ Ulll.,t~.l for detection of patient posture and activit~
level, ~I Li.,ulA . ly to provide A~/t/IU, ' pacing rates during stair climbing and ~IPer~n(1in~
~PcAriptinn ~A,f the prinr Art Rate responsive pacing has been widely adopted for adjusting pacing rate to the pll~ logic needs of the patient in relatively recent years. Early single chamber cardiac provided a fixed rate stimulation pulse generator that could be reset, on demand, by sensed atrial or ventricular c. .,.~ ...c recurring at a rate above the fixed rate.
a o Later, dual chamber demand ~ became available for;,.~ ;. . in patients having an intact atrial sinus rate but no AV Aon~ Atir~n so that ventricular pacing could be ~yll~,LIvlll~.d with the atrial sinus rate, and backup fixed rate ventricularpacing could be provided on failure to sense atrial ~lpp~ ri7~tione In addition, rate ~ , A' ~ le 1'~'` .",.k~, ~ became available wherein the base pacing rate could be selected by a 25 physician to provide a CUIIIIJ1UIII;.~ fixed rate that did not interfere with patient rest and provided adequate cardiac output at moderate levels of exercise Such fixed rate pacing, pAlLi~.UiAlly for patients not having an adequate atrial sinus rate to allow syncluullvu~ pacing, left most patients without tne ability to exercise, lift objects or even walk up stairs without suffering loss of breath due to i "~. . lTi~i~ . ,l cardiac 3 o output. However, the ;llLIudu~,~;u~l of the M~d~lU.L~ Activitrax~ pacemaker provided patients with the a pulse generator having a rate responsive capability dependent on the level of patient activity. A l.~ . crystal bonded to the interior of the ;. . .I,l , .: _1,1.
pulse generator can or case is employed in that pacemaker and successor models to provide a pulse output signal related to the pressure wave generated by a patient's footfall wo 96/30079 ~ , ; . 2 ~ 4 and conducted through the body to the crystal. Thus, low frequency activity signals recurring at the patient's rate of waLking or running could be sensed and processed to derive a pacing rate appropriate to the level of activity. The activity sensor and its operation is described in commonly assigned U.S. Patent No. 4,428,378 to Anderson.
Since the introduction of the Activitrax~ pacemaker, a great many rate responsive employing a wide variety of activity sensors and other physiologic sensors have been proposed and marketed. A Cu~ u~ ;vC listing of such rate responsive rA, . .11A~I~rl ~ sensors and sensed physiologic parameters is set for~ in commonly assigned U.S. Patent No. 5,226,413 to Bennett et al., illcul~ul~L~d herein by reference. However, 0 the activity sensor of the type employed in the Activitrax~ pacemaker continues to be used in successor single and dual chamber, rate responsive pacemaker models and remains the most widely used physiologic sensor.
As mentioned above, this ~ crystal sensor is responsive to pressure waves generated by patient footfalls striking the exterior of the pulse generator case.
Activity sensor A~A,nfiALl~rAtiAIn~A employing integrated circuit, AC dcc~ u~ t~la on an IC
chip inside the pacemaker are also being employed in the EXCEL"VR pacemaker sold by Cardiac ra~ 6cl~, Inc., and in similar rate responsive ~ sold by other ura.Luucl~. The AC dl,c~lclulllc;L~l is formed of a silicon beam mass suspended on the IC that swings or moves in response to shock waves caused by body motion and provides 2 o an output signal having a magnitude dependent on the rate of movement.
The relative virtues and ~ . ' of l.~ , ;. crystal and AC à~
activity sensors and associated pAr~mAk~Ar.c are reported widely, e.g. in the article "Activity-Based Pacing: (`omrAAri~A~An of a Device Using an A~ lu~ t~l VerSus a r~ " '~ Crystal", by Bacharach et al. (I~E, Vol. 15, pp.l88-196, February 1992).As indicated in that article, the pacing rate responses of these pAr~nn~ rA~ strapped on patients with normal hearts who were subjected to various stress tests were measured and compared to each other and to the patients' average actual heart rates. The tests conducted included stair ascending or climbing and descending tests, and ~ '!'' Il'`i''ll` were drawn to the effect that the AC â~h,lulll~t~l performed superiorly to the ~ uclccLI;c sensor in 3 o certain respects. Higher cardiac output is required in ascending a flight of stairs than in walking at the same rate or in descending the flight of stairs at the same rate as indicated by the patients' heart rates. The reported AC dC.~I.lulll.,..l induced pacing rate during stair climbing more closely matched the required cardiac output as indicated by the test .. . . .. ..... ..... . . ... . _ . .

WO 96/30079 t,~ t S 3 2 1 9 0 1 5 9 PcTlus96lo232~
subjects' average heart rates. During stair ~lPccf n~in~ the AC d.,~,CI~.V~ induced pacing rate did not a,u~ ,;dbl~ fall and exceeded the patients' actual heart rate. The reported ~ f Ic~ sensor induced pacing rate during stair climbing fell below therequired cardiac output as indicated by the test subjects' average heart rates. During stair 5 t1f ~nf n~in~t the ~ f ~ crystal induced pacing rate increased from the rate achieved during ascending and also exceeded the patients' heart rate.
As a result, while the authors suggest that the AC dCC~ ulll~t~,l is superior incertain respects to the ~ f ~ l, ;c crystal sensor, the test data also indicates that the AC
ac. c,l~lu..l.,.,,.a do not adequately distinguish between stair ascending and descending or 0 walking at the same rate on a flat surface to set an appropriate pâcing rate. Neither the AC
d~ ,lulll~t~,l nor the ~ uel~ ;c sensor can inherently distinguish these pâtientactivities. If an appropriate rate for an imdividual patient is set for stair climbing, for example, that rate may only be triggered by the frequency of recurrence of the patient footfalls and ..., .~ ly may be too high a rate for either stair descending or level 5 walking at the same speed.
Like the ~ f ~ ;r crystal sensor~ there is no signal output from the Ac accelerometer in the absence of body motion and related to body position or attitude. In other words, when a patient is at rest, neither activity sensor provides any indication as to whether the patient is upright and awake and resting or Iying down and presumabl~
2 o sleeping or resting. Other sensors for sensing physiologic parameters induced by high levels of exercise have been proposed to detect the physiologic changes a~cul--~exercise, rest and sleep to trigger appropriate rates. To lower the pacing rate during sleep, the inclusion of a real time clock to establish a Circadian rhythm pacing rate has also been proposed. None of these proposed sensors or systems are capable of fif tf rminin~ a 2 5 patient's position or posture.
A mf -~h~nir~l sensor has been proposed in the a~fticle "A New Mf ~h:~ni~l Sensor for Detecting Body Activity and Posture, Suitable for Rate Responsive Pacing" b~ Alt et al. (~E, Vol. I l, pp. 1875-81, November, 1988, Part 11) and in U.S. Patent No.

4,846,195 that involves use of a multi-contact, tilt switch. This switch employs a mercury 3 o ball within a container that is proposed to be fixed in the pulse generator case, so that if the pulse generator is implanted at a certain orientation, and stays in that orientation, certain contacts are closed by the mercury ball when the patient is upright and others are closed or none are closed when the patient is prostrate, i.e.. either prone or supine. During w096/30079 ~ ~ ~, 4 2 1 9 0 1 59 r~ 74 movement of the body, the mercury ball is expected to jiggle randomly and the number of contacts made per unit of time may be used as a measure ûf the level of activity. Similar sensors have beenproposed in U. S. PatentNos. 4,869,251, 5,010,893, 5,031,618 and 5,233,984.
The use of elemental mercury is generally not favored and would incrçase Cll ~dlull~ lLrl problems related to disposal of the pulse generators after use. Long term contact ~""l~ ;..,. and bridging issues would also arise, palLi~ulally given theextremely small size ofthe switch for ~i.,.l;"- .". .1l within modern pulse generator cases.
Plc~ul~l~ly, the multi-contact tilt switch sensor would also not necessarily be able 10 to distinguish between stair climbing and descending at the same stepping rate. Given the necessary small size of the tilt switch, it would be difficult to accurately position the pacemaker pulse generator so that consistent, reproducible signal outputs from the sets of contacts bridged whilç stûoped forward or rearward would be achieved in a given patient over time. Moreover, the limited nulnber of contacts reduce the possibility that such 5 fli~ ;. ." could be achieved. To date, no implants of pacemaker pulse generators using such a tilt switch have been reported.
More recently, the use of a solid state position sensor in the form of a DC
acc~ "ulll..,.l is proposed in U.S. Patent No. 5,354,317. The DC a~,~,cl, lulllctcr disclosed in the `317 patent is fabricated in hybrid ~ IC form as a polycrystalline 2 û silicon, square plate, suspended at its four corners above a well in a single silicon crystal substrate, and associated low pass filter circuits are formed on the same substrate. The suspended plate structure moves between stationary positions with respect to the well on the suspension arms in response to earth gravity, depending on its orientation to the gravitational field. The plate also vibrates on the suspension arms similar to the AC
2 5 ~c~ lul,l~ in response to arr~lrr~tirn movements of the patient's body.
The single DC ac~cl..ulll~t~, of the '317 patent is oriented to be sensitive to the anterior-posterior axis of the patient so that the upright, supine and prone body positions can be ~icrrimin:~tr~1, and separate base pacing rates can be set. Rate changes from the base pacing rates dependent on the exercise level of the patient in each position are 3 o suggested. When changes in Fatient position are detected in the absence of physical exercise, the base pacing rate change is smoothed between the old and new rate to avoid a sudden step change.

wo96/30079 i ~ 5 2 1 9 1~ 1 59 r l/U~., 61A~t The signal processirlg of the output signal from ~he single DC acc.,lelulll~t.l of the `317 patent includes signdl level calibration for each individudl patient to account for differences irl the angle of orientation of the DC æ~,cL,. o~ plate resulting from the angle of the pulse generdtor case in the patient~s body~ However, this 5 calibration is not suggested in order to distinguish body positions hdving a more or less commonangularrelationofthemovableplatetothe~av~ uldlfield.
Irl addition, the ' 317 patent does not appear to suggest any .1; ~ of stair climbing that would alleviate the problems identified above resulting in the same or a higher pacing rate being developed during stair descending than during stdir climbing.
Despite the ~ reported with respect to the p;~ elc~,tlic sensors and solid stdte æccl.,~ullletcrs, they remain favûred over the other physiologic sensors that have been proposed or are in clinical use due to their relative simplicity, reliability, predictdbility, size, and low cost.
Probll mc tn be Snlved by th~ Jnventinn In view of the 'J ' ' ~ advantages of the ~ and AC dCCCI~ t~
type activity sensors, it would be desirable to employ solid stdte sensors responsive to patient activity in a similar manner tbdt would also distinguish stdir or steep incline climbing from other activities in order to provide an appropriate rate response to provide adequate cardiac output.
2 0 SUMMARY OF TH~ INVENTION
In view of the above, it is an object of the present invention to provide a rateresponsive pacemaker employing a body position sensor to distinguish stair climbing from other activities, e.g. stair descending or walking on a level surface, and to provide an .:LI.I~)I I.l~JI ;dtt~ pacirlg rate increase from a rest rate during stdir climbing.
2 5 It is yet a fuîther particular object of the present invention to provide such pacing rate setting capabilities to provide a higher pacing rdte for a pdtient ascending stdirs or a steep incline than descending stairs or walking on a relatively level surfæe.
Tbere is provided in accordance with the present invention, a rdte responsive pacemaker for pacing a patient's heaît at a pacing rate dependent on patient ætivity and 3 0 poSture pa~ LU Iy during stdir climbing, at least including the means for and steps of:
deriving a bod~ posture tilt signal having a .1.~ ;. varying ~vith the degree towhich the patient posture is in an upright stdnce or leaning forward;
detecting patient footsteps;
, w096/30079 j~3~ , 6 ~ I 9 0 1 59 r~ 4 deriving a patient activity signal havmg a signal level dependent on the frequency--of patient footsteps recurrirlg over a time unit;
deriving a rate control signal from the bod~ posture tilt signal arld the patient activity signal correlated to the physiologic demand on the patient's heart;
defining physiologic escape intervals as a function of the rate control signal to establish a pl.~.;olo~ic pacing rate;
generating pacing pulses at the physiologic pacing rate; arld applying the pacing pulses to the patient's heart.
Preferably, the posture of the patient is determined through the use of a solid state, DC a.,ccl~lull.~ mourlted within the pacemaker pulse generator case having a sensitive axis aligned with the pacemaker case and the patient's anterior-posterior (A-P) body axis.
The DC acccl~,lul.l~ provides an output signal due to the force of gravity which has a polarity and magnitude dependent on the degree to which the sensitive axis is tilted forward or rearward from the direction of earth's gravity. Forward lean or tilt of the patient while upright ~ ". .~ 1 by a recurring series of footfalls can be ~
from an upright stance ar~d a similar level of footfalls to thereby distinguish stair climbing from other activities in the same stepping rate range and provide an appropriate pacing rate for each activity.
llle DC acc~l~,lv~ t~l is preferably mounted into an IC chip with a second and 2 o optionally a third DC acccl~,lul.l.,tl so that their sensitive axes are aligned with the three axes of the pulse generator case. The physician can implant and stabilize the pulse generator case ~n the proper orientation to the patient's thorax to align the sensitive axes with the superior-inferior (S-l), anterior-posterior (A-P) and lateral-medial (L-M) axes of the chest cavity. As a result, distinctive signal levels are developed by each DC
2 ~ acc~l~,lulllctcr in each posture position due to the effect of gravity on the sensitive axis of each gPn irnn~ t~r element. From these signal levels, the posture of the patient can be determined for providing additional pacing rates appropriate to the other determined body positions and the activity level of the patient.
~d~ulLab~uuDl~/, one or more ofthe DC a~.L,.ulll~ D can be used to derive the 3 o level of patient aclivity from the number of changes in signal levels exceeding a certain threshold occurring in a given sampling time period. as is conventional in use of the c and AC accch"ul,lct~,l activity sensors described above.

w096/30079 ~ 7 2 1 9~ 1 59 r~ 4 The present invention may also be ;~ ,t I employing other for ns of body position or tilt sensors having a sensitive axis in the A-P direction, p~uliu~l~ly the sensor disclosed in the above-referenced `984 patent.
It should be noted that the DC acc.,l~,lul..~,t~,. of the above-referenced `317 patent is 5 a bulk ~lliululll~,lf~ ,d IC structure thdt has a serlsitive axis normal to t_e plane of the movable plate and provides the +l, - l and O static output signal levels depending on the orientation of the sensitive axis to the vertical grdvitational force. If such a DC
a. ~ lu~ t~l is used in the practice of the present invention, the l-rth~-.ns~lly arranged DC dC~ lulll~t~l~ would provide similar signal responses as long as the sensitive axes are 10 oriented in the sarne manner as described above~
Allv~nt~e~ Qf the Tnventi~-n The DC output signal of a DC a~ c~l~lull..,i.l can be processed to detect body for~vard tilt, while the patient moves at a walking pace, and thereby employed to .l;~., i,, ., -i . ~. stair climbing from other activities and to develop an a~lul ' ' pacing rate, 15 solving the problems associated with the prior art rate responsive ~ PI ~ employing activity sensors. I~he DC ~CCCI~.U~ ,l and associated circuitry can be easily illl ullJvla~d into a pacemaker pulse generator at low cost. The eæe of use, and the reproducibility and consistency of results attained will lead to ~ pt~hility within the medical community.
BRIEF DE~CRTPTION OF TTTF. DRAWINGS
2 o These and other objects, advantages and features of the present invention will be more readily understood from the following detdiled description of the preferredemh~1im~-nt~ thereof, when considered in conjunction with the drawings, in which like reference numerals indicate identical structures throughout the several views, and wherein:
Figure l is block level diagrdm of a DDDR pacemaker capable of i., .l,l ., ~ at 2 5 least one of three possible, mutudlly orthogonal DC a. ccl~l Ulll~tC. ~ as activity and patient posture sensors ~ l~ly to detect forward tilt;
Figure 2 is a sc~ematic illustration of the . ~ of the S-l, L-M, and A-P
sensitive axes of three DC .ICC-,lClv~ ,t~,.'i mounted orthogonally with respect to a hybrid circuit substrate mounted within the housing for the pulse generator of Figure l and the 3 o markings on the housing for orienting the pulse generdtor with the patient body axes;
Figure 3 is a rate response overview flowchart of the algorithm illcullJul.llPd into the pacemaker of Figure I for deriving a physiologic pacing rate related to stair climbing wo 96/300 ~ 3 ~ J 8 ~ 1 q U 1 5 9 ~ ~ ~4 from the output signal of the DC al~ ,lUll.~.t~.l of Figure 2 oriented alorlg its sensitive axis--in the A-P direction:
Figure 4 is a detailed flowchart of the stair climbing .l;~, ;, . .;, . -~ i nn step of the flowchart of Figure 3;
Figure 5 is a detailed flowchart of the lli~ ;.. ,;., ~; ") rate calculation step ofthe flowchart of Figure 3;
Figure 6 is a graph illustrating the calculation of the appropriate pacing ratesrelated to the degree of body tilt of an active patient in walking, climbing and descending a flight of stairs;
Figures 7-9 are graphs illustrating the tilt deviation ~ u~ resulting from of tests conducted on test subjects employing the stair climbing li~ , algorithm ofFigures 3-5; and Figure 10 is a graph illustrating the delivery of the appropriate pacing rates related to the degree of body tilt of an active patient in walking, climbing and descending a flight 15 of stairs.
DETATT Fn DF~CRTPTION OF THF PRF.FFRRFT~ T~MRODIMF~TS
The present invention is preferably i~ , d in multi-iu~ "~ r DDDR
IJ " . .1~ I rl ~ of types widely known in the prior art. However, the invention could be imrlPnnPn~Pd in simpler, single chamber ~ rl ~ As described above with respect to 2 o other medical devices, the irlvention may also be i., ,l .l,., . ~l in other medical devices for providing other therapies and or for monitoring physiologic paramekrs in the various body positions the patient may assume where stair climbing .li~ ,,, may be important.
Figure I is block level diagram of such a pacemaker il"l,l~ pulse generator or 2 5 IPG 30 and lead set 12 and 1~ which sets forth the structures required to incorporate the invention into a DDDR pacemaker. In the drawing, the patient's heart 10 has an atrial pacing lead 12 passed into the right atrium and a ventricular lead 14 passed into the right ventricle. The atrial lead 12 has an atrial electrode array 16 which couples the pulse generator 30 to the atrium. The ventricular lead 14 has a ventricular electrode array 18 for 3 o coupling the pulse generator 30 to the ventricle of the patient's heart 10. Atrial and ventricular leads 12 and 14 are depicted as bipolar leads coupled to a bipolar IPG 30, although unipolar leads could be employed with a suitable IPG.

WO 96/3007s ~ !i t~ 2 1 9 0 1 ~ 9 ~ 7~-4 The IPG circuit 30 of Figure I is divided generally into a pacing circuit 32 coupled to a battery power supply 50, an activity sensor 60 of the type described below, a telemetry coil 45 and a ~ ,lùCUul,uuttl circuit 34. The pacing circuit 32 includes the atrial and ventricular output amplifier circuit 36 and sense amplifiers 38 that are coupled to the atrial and ventricular leads 12 and 14, I~ ,.,Li~.ly, the digital controller/timer circuit 40 and other associated ~ described below. The output circuit 36 and sense amplifier circuit 3 8 may contain atrial and ventricular pulse generators and sense amplifiers ~,ull~ulllillg to an~ ofthose presently employed in Cullll~ ,;dlly marketed dual chamber cardiac u~
Sensed atrial ,1. I,~,l,.,;, ,. l ;. " .s (A-SENSE) o} P-waves that are confirmed by the atrial sense amplifier are ~."...,..".;. - ' 1 to the digital controller/timer circuit 40 on the ASE line. Similarly! ventricular d ~ (V-SENSE) or R-waves that are confirmed by the ventricular sense amplifier are ~ . ,.,..,..,..: ..~ to the digital controller/timer circuit 40 on VSE. The sensitivity control block 42 adjusts sensitivity of each sense amplifier in response to control signals provided by digital controller/timer 40 that are in turn stored in memory in Ill;~,lU~,UIII~ circuit 34.
In order to trigger generation of a ventricular pacing or VPE pulse, digital controller/timer circuit 40 generates a trigger signal on the V-trig line. Similarly, in order to trigger an atrial pacing or APE pulse, digital controller/timer circuit 40 generates a 2 Q trigger pulse on A-trig line.
Crystal oscillator circuit 44 provides the basic timing clock for the pacing circuit 30, while battery 50 provides power. Reference mode circuit 48 generates stable voltage reference and current levels for the analog circuits within the pacing circuit 30 from the battery voltage and current. Power-on-reset circuit 46 responds to initial connection of the circuit 30 to the batter~ 50 for defirling am initial operating condition and also resets the operating condition in response to detection of a low battery energy condition. Analog to digital converter (ADC) and m-lltirlr Yflr circuit 52 digitizes analog signals and voltage to provide real time telemetry of ASE and VSE cardiac signals from sense amplifiers 38, for uplink trAncnniCCi~-n ~ ia RF tratlsmitter and receiver circuit 47. Voltage reference and bias 3 o circuit 48, ADC and I ~ Jr~ 52, power-on-reset circuit 46 and crystal oscillator circuit 44 may correspond to any of those presently used in current marketed illlul~lt blc cardiac wo 96/30079 ~ i T '-- 1 5 9 r~ . 7'~'~4 Data ~,",.~.,..~;..,. to and from an external ~UIU~ lUll~,l (not shown) is s~ c. .. "~ l .. ;l by means of the telemetry antenna 45 and the associated RF tramsmitter and receiver 47, which serves both to tl~-mo~ lot~ received downlink telemetry and to transmit uplink telemetry. For example, circuitry for ~ mr\~ otin~ and decoding downlink telemetry may correspond to that disclosed irl U.S. Patent No. 4,556,063 issued to Thompson et al. and U.S. Patent No. 4,257,423 issued to McDonald et al., while uplink telemetry functions may be provided according to U.S. Patent No. 5,127,404 issued to Wyborny et al. and U.S. Patent No. 4,374,3~2 issued to Markowitz. Uplink telemetry capabilities will typically include the ability to transmit stored digital inf~rm~tion as well 1 û as real time or stored EGMs of atrial and/o} ventricular electrical activity (according to the teaching of the above-cited Wyborny patent), as well as 1. ~Ula~ a;Vll of Marker Channel pulses indicating the occurrence of sensed and paced ~1~p~1Orj7~tionc in the atrium and ventricle, as disclosed in the cited Markowitz patent.
Control of timing and other functions within the pacing circuit 30 is provided by di~ital controller/timer circuit 40 which includes a set of timers and associated logic circuits cormected with the Illil,lU.,Ol~l,UU.~I 34. Mi~,lu~ulllu~lL~l 34 controls the operational functions ûf digital controller/ timer 40, specifying which timing intervals are employed, and cûntrolling the duration of the various timing intervals, via data and control bus 56.
Mi~luculuuuL~I 34 contajns a Illi~luAuluc~.~aui 54, associated system clock 58, and on-2 û processor RAM and ROM chips 64 and 66, l~,aAu~ i v~ly. In addition, Illiuluculll~uu~er circuit 34 includes a separate RAM/ROM chip 68 to provide additional memor~ capacity.
Mi~,lu,uluu~,~aul 54 is interrupt driven, operating in a reduced power consumption mode normally, and awakened in response to defined interrupt events, which may include the A-trig, V-trig, ASE and VSE signals. The specific values of the intervals defined are controlled b~ the IlliWULUlllUU~I circuit 54 by means of data and control bus 56 from ,ulu~l<uuuu~l-in parameter values and operating modes.
If the IPG is ~lu~l~ulullcl to a rate responsi~ e mode, tbe patient's activity level is monitored periodically, and the a sensor derived pacing escape interval is adjusted ,ulu,uulliull~ . A timed interrupt, e.g., every two seconds, may be provided in order to 3 o allow the mi-lul,lu~ aul 54 to analyze the output of the activity circuit (PAS) 62 and update the basic V-A escape interval employed in the pacing cycle. In the DDDR mode, the V-A escape interval may be selected as the variable pacing rate ~lhl ,l;~l,;l,~ interval, W0 96/3~079 ~ , ~ ? ~ 9 0 l 5 9 1 ~ l",~. ~ 7~74 but the A-V interval and the atrial and ventricular refractory periods may also vary with the V-A escape interval established in response to patient activity.
Preferably, two separate lower rate V-A irlterval timer functions are provided. The first is set by the physician when the base pacing rate is selected. This DDD V-A time 5 interval starts from the occurrence of a VPE or VPE, and provided neither an ASE nor a VSE occurs during the V-A time interval, an APE is generated after the expiration of the V-A time interval. The duration of the second lower rate time interval is a function of the measured patient activity acquired by the activity sensor 21. Typically, this DDDR, V-A
time interval begins with a VSE or VPE and has a time duration reflecting patient activity.
10 In this art, such structures are well kno~hn, and a variety of techniques can be used to irnplement the required timer functions.
Digital controller/timer circuit 40 starts and times out these and other intervals employed over a pacing cycle comprising a successive A-V and V-A interval in a manner well known in the art. Typically, digital controller/timer circuit 40 defines an atrial 5 blanking interval following delivery of an atrial pacing pulse. during which atrial sensing is disabled, as well as ventricular blanking intervals following atrial and ventricular pacing pulse delivery, during which ventricular sensing is disabled. Digital controller/timer circuit 40 also defines the atrial refractory period (ARP) during which atrial sensing is disabled or the ASE is ignored for the purpose of resetting the V-A escape interval. The 2 o ARP extends from the beginning of the A-V interval following either an ASE or an A-tri~
and until a lu.~ d time following sensing of a ventricular ~lPpsl~-i7~ti~n or triggering the delivery of a VPE pulse. A post-ventricular atrial refractory period (PVARP) is also defined following delivery of a VPE pulse. The durations of the ARP, PVARP and VRP may also be selected as a ~)IU~Iallllll~bl~. parameter stored in the 25 Illi~,lUCUlll,UUL~ 34. Digital controller/timer circuit 40 also controls the pulse widths of the APE and VPE pacing pulses and the sensitivity settings of the sense arnplifiers 3 8 by means of sensitivity control 42. Digital controller timer/logic circuit 40 also times out an upper rate limit interval (URL) set by a value ~)IU~i~lllllled into memory in Ill;~lUCulll,uu~-circuit 34. This timer is initiated by the occurrence of a VPE or VSE, and limits the upper 3 0 rate at which ventricular stimuli are delivered to the heart. The lower pacing rate is eslablished by a ulUKI~IIIll~-in V-A or A-A interval value stored in memory in lli~lu~,uL,J~u~er circuit 34.

wo 96/30079 P !:~ i '3 ~. 12 2 1 9 0 1 5 9 T~,.l.J., -"'7~74 The illustrated IPG block diagram of Figure I is merely exemplary, and ~;UIII,~U~ to the general functional ul~ iull of most multi-ulu~ r.
Illil,lU~JIU~.C.~UI controlled DDDR cardiac p ~ presently cull~ lly available.
It is believed that the present invention is most readily practiced in the context of such a 5 device, and that the present invention can therefore readily be practiced using the basic hardware of existing III;~,IU,UlUl.,~ lUi controlled dual chamber ~ , as presently available, with the invention . ' ' primarily by means of mn~iifi.~tinnc io the software stored in the ROM 66 ofthe llli~lU~Ulll~JUt~l circuit 34. However, the present invention may also be usefully practiced by means of a full custom integrated circuit, for 10 example, a crrcuit taking the form of a state machine as set forth in the above-cited ~etzold et al. patent, in which a state counter serves to control an arithmetic logic unit to perform . .1. . .1,.1 ;. .1-~ according to a prescribed sequence of coumter controlled steps. As such, the present invention should not be understood to be limited to a pacemaker having an nl ~llit.~ Lu.~ as illustrated in Figure I .
Figure 2 is a schematic illustration of .Illbo luu~,uL of a DC a.,c~l~,lulll.,.. l based forward lean sensor that may be employed in the practice of the present invention. In Figure 2, three solid state, DC accel~lulll~tcrs, namely the S-l DC ac~,cl~lulll.,L~l 72, A-P
DC iiCC~IClulll~ l 74, and L-M DC dCC~I.lUlll~..,l 76, are mounted so that their sensitive axes are ortbogonally directed to the S-I, A-P and L-M axes, I ~ iV~Iy, of the pulse 2 o generator hybrid circuit substrate 76 and exterior case 78. In the practice of the present invention, the DC output signal of the A-P DC accelerometer 74 is preferably employed in the .I;~. .;,.,i,.~lion of stairclimbing from otheractivities.
Each of the DC ~cccl~lu~ 72, 74, 76 is preferably a surface Illi~lullla.llill~.lintegrated circuit with signal . ..~,.I;li..,.:.,g, e.g. the Model ADXL 50 dcc~h,.ulll~t~l sold by 25 Analog Devices, Inc., Norwood MA and described in detail in the article "Airbags Boom WhenICAccc;l~lulll~t~,lSees50G",intheAugust8,1991,issueofFlPrtroni(~I~)P~i~n andin "Monolithic Accelerometer with Signal Cnn~itinnin~", Rev. O, published by Analog Devices, Inc., both illCUI~ul~L~1 herein by reference in their entirety. Employing surface llu~l"",~.I,;";"~,asetofmovablecapacitorplatesareformedextendinginapatternfrom 3 0 a shaped polysilicon proof mass suspended by tethers with respect to a further set of fixed polysilicon capacitor plates. The proof mass has a sensitive axis along which a force between OG and +/- 50G effects physical movement of the proof mass and a change in Wo96f30079 ~-'?f"?~ 3 ? 190159 ~ u~ 7~
measured ~ , e between the fixed and movable plates. The measured ~ A ; 1~ ' 1' C is r( ~ by the on-chip signa~ circuits into a low voltage signal~
The proof mass of the ADXL 50 is co-plamar with the IC chip plane it is tethered to for movement back and forth in positive and negative vector directions along a single 5 sensitive axis. The planar orientation thus provides that the proof mass sensitive axis is along the length of the proof mass. For offthe shelf use, the ADXL 50 IC chip is mounted in a TO-S can with the positive vector direction of the sensitive axis a igned to a reference tab of the can. By using to the can tab, the positive or negative vector direction of the serlsitive axis cam be aligned with respect to some plane or angle of the system or circuit it 0 iS used in with respect to the constant vertical direction of ~laviLaLiullal force.
The reference tabs for the three axes are ~. I .. . "Af ;1 ,.11y illustrated in activity sensor 60 of Figure I and with respect to each of the DC d~ u~ Lcla 72, 74 and 76 of Figure 2. Of course, in actual custom faWcation within the pulse generator 30, the DC
acc~ lulll~t~la would be formed or assembled on a single IC cbip and the assembly could 5 be enclosed in a single IC package moumted to hybrid substrate 60. The assembly of the hybrid substrate 76 within the pulse generator case is precisely controlled to establish the ~lrientAtir,n The S-l, A-P, and L-M orientation markings X2, 84, and 86 may be made on the pulse generator case ~8 for the ~.UII~ .Ci of the implanting physician.
The effect of I G of gravitationa'f force applied directly along the sensitive axis of a 2 0 stationary ADXL 50 d~,1elo--l~,t~,l provides a ~ ;c output voltage signal level that is referenced or scaled as +I for angular c. mr~tAti~-n purposes. The effect of I G of gravitational force applied in precisely the opposite or negati-~e direction to the sensitive axis provides a ~hArArtPri~tir output voltage signal level that is referenced or scaled as -I .
If thesensitiveaxisisorientedLl~la~la~l~ytothedirectionofthe~l~v~ ivl~dlforce~a 25 bias voltage level output signal should be present, and that voltage signal level is referenced or sca ed as 0. The degree to which the sensitive axis is oriented away or tilted from the direction of the ~-aviL~lliulldl force can also be detected by the magnitude and polarity of the output ~ oltage signal level deviating from the bias level scaled to 0 and below the outrut signal level values scaled to ~1 and -I . The above-referenced 3 o publications provide illallll~,liulla for sca'fing the voltage signal levels to the û, +l and -I
static level values. A Illiwulllv~ aul interface circuit with auto calibration of offset error and drift caused by t.~ variation that may be employed in the activity circuit 62 of Figure I is also described.
.

W096/30079 ~?~ 14 219~159 ~
Other scales may be employed, depending on the signal polarities and ranges employed. The examples described below with reference to the datd collected in testing and illustrated in Figures 7-9 employ a scale where OG develops a +1.000 volt DC signal, +IG develops a +1.400 volt DC signal and -IG develops a +0.600 volt signal.
The effect of i, .~ ,. v~ ~ or AC charlges due to body motion q- ~ . l. . ,.1 ;. ." can be measured by the voltage signal output level changes per unit time. As indicated in the above-iul.,ul,uul ' l."l,li. ~ ,.,c the ADXL SO can, ~ "- v ~
~ rf~ r,qti~n levels up to 50Gs, which is well in excess of the sensitivity required to detect patient footfalls regardless of the intensity level that a patient could muster. The output signal levels may be scaled to a lower rarlge, e.g. O to +2-SG through adjustment of the internal ADXL 50 buffer amplifier or custom fabrication.
Returning to Figure 2, the present invention may be illculluulaL~d in an IPG having a single one, two or three DC dC.,CI~.ul.~t~l~, the selection of a single one at least including the A-P or S-I DC a.,ccl.,.ulllct~ 74 or 72, lc~ ,ly, and preferably the A-P
DC dcccl~lulllctcr 74. Figure 2 thus inclusively illustrates any such comhinqti~n and the following description of the c. mhinqti~-n of all three will be understood to be inclusive of lessthanthreeforpurposesof,...fl~.~l~,..l;,.~thepresentinvention. Ofcourse,thepresent inventiûn may be adv~lLageuu ,ly combined with the system for l1rlr~ e other body positions than stair climbing employing the output signals of the other d.~ ullleters in 20 ~nnhinqtinn When the one, two or three DC ac~ ulll~ 72, 74 or 76 of the ADXL 50 tS~pe are incorporated into a pulse generator as depicted, the sensitive axis of S-l DC
accck,lulll~t.l 72 is intended to be aligned, when the pulse generator 30 is implanted, as close to vertical as possible, employing the markings B2, 84, 86, for example. Thus, ~hen 2~ standing upright and remaining still, the output signal level generated by +lG should be realized or closely approached by the S-I DC dC.,.I~lulll~t.l 72. At the same time, the output signal levels of the A-P and L-M DC d.,..cl~lulllcters 74 and 76 should approach those l~ llLil~g OG.
When the patient lies still on his her back or stomach, the DC signal levels of the 3 o A-P DC ac.cl~,lvlll~t~l 74 should approach those generated by +IG or -IG, respectively, (if the pulse Eenerator case 70 is implanted with the A-P DC ac~lclulll~,t~. positive vector pointed anteriorly) while the signal levels of the S-l and L-M DC a.,~ .ull..,i~l~ 72 and 76 should approach the DC signal level generated in response to OG. In the same fashion, , .,:: .. :, ,, , , , .. , , ., , , .,, , , . . , ., _, _,, , . , _ ,, ,,, _ .. , : _ , _ _ _ .

w096/30~79 ~ f'-r~ 901 59 r -~u~ 7~74 the patient Iying on the right and left sides will orient the sensitive axis of the L-M DC
C~ ull..,t.l 76 with the gravitational force to develop either the + I G or -1 G signal level while the signal levels of the S-I and A-P DC ~ CCCI.lu~ ,t~ 72 and 74 shûuld approach the OG signal level.
Deviations from the DC signal levels . I, - ,.. t- . ;`l ;~' of + I G, OG and - I G of each DC d~ ,lUI~ 72, 74 and 76 can be measured after impl~nt~til~n during a patient work up in these positions. The deviations may be stored in RAM 64 as adjustment values to be used by the Inl.~lu~lùc~ ul in weighting or otherwise processing the actual scaled output signal levels of the three DC al,,,.l.,.ul,l~ 72, 74 and 76 periodically supplied to the 0 Ulh l UCulllLluL~l circuit 34 through the digital controller/timer circuit 40. Moreover, the actual il~ luu~lLiull orirnt~ti~ n~ of the positive axis vectors of A-P and L-M DC
acccl~.u.ll~t~.~ 74 and 76 can also be determined by the polarity of the signals generated.
Those orientations may be stored in the l"i~,ucuu,,uui.. memory and employed to chamge the polarity of the output signal levels of the three DC aL~,.,l~.u",~,t~.~ 72, 74 and 76, as 15 necessary.
The above description provides a framework for developing a set of equations forderiving the patient's physical position while at rest and while moving in a variety of positions as described in greater detail in the above-referenced '(P-327QB! application. In accordance with the present invention, less than three orthogonally disposed DC
2 o acc.l~,.u.,l~,t..~ of the type described above may be employed in stair climbing ~ . .;", i" ~ n Preferably, the output signal of the A-P DC a- c~ ullleter 74 in the range between O and +IG (depending on the orientation of gravity to the sensitive axis) and the frequency of 0-0.5 Hz is detected and employed to determine if the moving patient is leaning forward and the extent of forward lean or tilt. The A-P DC ~cc~l..ul"~t.. 74 also 2 5 generates recurring AC :~rr~-lrrslti~ln output signals in the same or higher magnitude range and a frequency of I - l O Hz indicative of footsteps or other body motion. The combination of signals is employed in the ,1;~. . ;",;., -~;. ", of stair or steep incline ascending by of a "Tilt Deviation" signal level to thresholds as described below. Simple bending over motion or the static forward lean of the patient in the absence of the recurring 3 o signals exceeding a threshold activity level is not determined to be stair climbing.
Similarly, too great a forward lean or tilt ~2 ~"",~ ,1 by activity signals exceeding the threshold will not cause the stair climbing heart rate to be invoked.

wo 96l30079 ~ t ~ 16 2 1 9 ~ 1 5 q PcrlUS96102324 Turning to Figure 3, it depicts a rate response overview flowchart of the algorithm ill~,UII~ul~.~d into the pacemaker of Figure I for deriving a ~I. y ,;ulo~i~; "Stair Climbing"
pacing rate from the output signal of the AP DC a~ccl~,lv...~.~l 74. Figure 4 is a detailed flowchart of the stair climbing (' step of the flowchart of Figure 3. Figure S is 5 a detailed flowchart of the di~l rate calculation step of the flowchart of Figure 3.
Figure 6 illustrates the selection of the Stair Climbing Rate or an ~ . ,r .l;-t pacing rate versus a Target Rate depending on the Tilt Deviation signal level as a Pacing Rate Control signal for use in setting the pacing rate (aRer any rate smoothing in the flowchart of Figure 3).
As described above, the A-P DC a.,cr~ u.llctcr 74 is oriented when the IPG case is implanted to the force of gravity so that the DC output signal level is nominal~y at +I .OOO
volts at OG and varies bet~veen +0.600 and +1.400 volts at -IG and +IG, respectively. In Figure 3, the signal output level from the A-P DC ac.,~ ull.ct~l 74 in block 200 is sampled at a sampling frequency, e.g. 200 Hz, and applied to block 202 where the DC
15 component "Measured Tilt" and the AC component "Activity Count" are determined over a certain sampling period, e.g. a running 2 second period. The Measured Tilt signal reflects both the DC signa~ level contributed by the forward lean or tilt of the patient's torso and the AC signal level that changes in magnitude as a function of the impact force from footsteps or other body motion. However, the AC signal . "A~ Ir; tend to be2 o averaged out the sampling time period. Certain of the Measured Tilt signal levels are averaged as described below. The current exercise activity level of the patient ma~
be derived from a count of the activity events. An activity event is detected in step 202 when an output signal of A-P DC d~,C~I~lulrl~t~l 74 (or one of the other DC acccl~lulll.t~
72 or 76, if present, or a ~ 1 of output signals) in the frequenc~ range of 1-10 Hz 2 5 is detected that exceeds a positive or negative scale threshold. The Activity Count is determined in a conventional process of filtering the sampled output signal in the 1- 10 Hz frequency range, amplifying the filtered signal, comparing the amplified signal to a threshold level, and coumting the threshold exceeding signals.
For example, the patient's footfalls cause shock waves to be transmitted through3 o the body that drive the A-P DC ~ I..ul~l~ter 74 to develop altemating output signals at a level exceeding the threshold level and within the specified frequency range for walking or rurming. Those sampled values exceeding the activity threshold level are ~ hli~-d as activity events. The activity events are counted in step 202 over a running time period, .. ... , . ... ... .. . , . . . _ . . _ . . . . _ . .. . _ . . .. . _ _ _ _ _ . , -WO 96130079 ~ 17 ~ 1 9 0 1 5 9 ~ 4 e.g 2 seconds, to derive the Activity Count. Arm and leg motion a~.,w~ g prone exercises, e.g. swimming, may also generate activity events.
In step 204, the Activity Count is employed to set the Target Rate appropriate to the estimated level of exercise. The Target Rate for pacing the patient's heart is 5 proportional to the Activity Count and varies bet~veen the ~ l.lllUllCid pacing Lower and Upper Rates in a manner well known in the art. Target Rate is typically used to refer to a pacing rate subject to further m--~ifir~fi~ n as by ~;ul~ Liolldl rate smoothing in physiologic l""~
Since the Activity Count for stair ascending may be equal to or less than that for 10 flat surface walking or stair ~r~n-~lin~, the Target Rate may be ;,,~ for stair ascending which requires greater cardiac output. Therefore, it is appropriate to empioy a substitute Stair Climbing Rate (again subject to mo~ifir~tinn before being used as the actual pacing rate). Since certain other activities may also generate an Activity Count mimicking walking or running, the .1~ 1~ . ",;" .l ;. ", of the posture of the patient with the 15 same or a set of the DC ~c~,~,L.lu...~t~l~ is important to a ~ t~rmin:~tirn that the Target Rate orthe Stair Climbing Rate (and i"t ",. .l;,,l r rates th~ . ) is ~
In accordance with the preferred ~ I.bo-li.ll~..~L, it is ~l."l~",~ l that a fixed Stair Clirnbing Rate may be programmed in for the individual patient. The Stair Climbing Rate is invoked if stair climbing is detected amd if the patient's activity level itself does not 2 o dictate a higher Target Rate. For example, the activity level due to the AC ~rr,~ rslti,~n component rr,ntrih~ltirm may signify light to heavy exercise levels dictating moderate to hi~h Target Rate of pacing. If the patient is moving rapidl~ ~ it may be immaterial that the patient is also climbing stairs or a steep grade, since the Target Rate may exceed the pre-l Stair Climbing rate in that instance. In a further variation where the Stair 25 Climbing Rate does exceed the Target Rate, the actual ;"t. . ",.~.-l; pacing rate may be selected to fall bet~-een the ~ d Stair Climbing Rate and the calculated Target Rate.
The stair climbing, l;~.,;, .,; " ~ ;.-" step 206 (shown in the flowcha~t of Figure 4) employs the Measured Tilt signal to make a fl~t~rmin~ltifm that the patient is or is not 3 o climbing stairs or an incline sufficiently steep and/or long to cause the patient to lean forward. The selection of the Stair Climbing Rate, the Target Rate or an ' rate for pacing the patient is determined in step 208 (shown in the flowchart of Figure 5).
Then, in step 2 l 0 of Figure 3, the pacing rate is modified to provide rate smoothing in wo 96/30079 ~ T ~ 18 2 1 9 o ~ ~ 9 PCTmS96/02324 f ~ from the lower Target Rate and the greater Stair Climbing or int~rm~r1is~tr rate--at the onset of stair climbing and ~ ;.,g back to the Target Rate on completion of stair climbing, employing well known rate smoothing techniques.
Turning to Figure 4, the stair climbing ~ .,. steps included in step 208 5 commence witb the ~ rmitl~tion of whether the patient is active in decision step 300 from the presence of an Activity Count. If no Activity Count is present, then the "not stair climbing" /1! trtmin~tinn is made or stated in step 302. No change in the Target Rate (in this case, the ,IJI`V~I ~IIIIII.~d Lower Rate~ is made, after the steps of the flowchart of Figure S
are completed.
1 v If the patient is active, then a Tilt Average is calculated in block 304 from the sum of the amplitudes of the Measured Tilt signals occurring while the Activity Count is satisfied (and as long as the Tilt DeYiation does not exceed an OTW Upper value as described below) divided by the number of samples. For example, a number of samples.
e.g. 300 samples, may be obtained on a running basis and i~.~,."",l .t. ~ on a FIFO basis 15 and the Tilt Average calculated therefrom. Typically, the Tilt Average will be derived when a patient is walking on a flat surface before climbing stairs and will reflect a DC
signallevelc.",;,;l.,.l;..-lonlytotheextentthatthelPGistiltedfromtheA-Paxis direction 84 at the imrl~nt~ti-~n site or the patient walks with a forward stoop. As mentioned above, any deviation due to inclination of the sensitive axis of the A-P DC
2 o a~ vl.l~t~. at the implant site may be determined during patient work up while the patient is standing still and upright. In the example described above, the obserYed deviation from ~1.000 volts may employed as a weighting factor to adjust the Tilt Average DC signal level component back to near the =1.000 volt level. However, in any given case, it may not be necessary to do the patient work up and make the adjustment.After a Tilt Average is calculated, a "Tilt Deviation" signal is then calculated in block 306 by subtracting the Measured Tilt signal from the Tilt Average. If there is no significant difference~ then the patient is continuing to walk or run without ~orward lean ~.I ,,.. ,.. I r~ of stair or steep incline ascending and the Tilt Average continues to be ac., ' 1. In this way, the onset of a DC component increase in the Measured Tilt3 0 signal attributable to patient forward lean can be detected by ...., ~ v,. to the Measured Tilt to the Average Tilt ~. ~ ...,...! .~ .1 while the patient remains active.
In step 308. the Tilt Deviation signal level is compared to an Outer Tilt Windov~
(OTW) previousl~ derived from the output signal of the A-P DC dc.,. l~lulu~"~l 74 during .. , .. . .. . . . .. . . . . .. ... . ... . . _ . .. .. , ..... _ _ , . . . ...

W0 96/30079 !~ 2 7 9 0 1 5 9 P~ 4 the previous patient work up. The OTW is an outer range of Tilt Deviations signal Yalues between an OTW Upper and OTW Lower value derived from the Tilt Deviation signalsgenerated in the work up as the patient ascends a set of steps or a stair step exercise machine for an average flight of stairs and is stored m the memory 68 of the micro-5 computer circuit 34. An Inmer Tilt Window (ITW) ~ Ulg a narrower range of theTilt Deviation signal values between an ITW Lower and an ITW Upper value may also be stored in memory 68. These ranges of values may alternatively be derived based on population studies and ~ ulull~,d by the physician. The relation of the OTW and ITW
in the selection of the pacing rate is explained further below in reference to Figures 5 and 6.
If the Tilt Deviation signal is not within the OTW, then it is determined in step 310 that the patient is not stair climbing, and the tilt deviation is checked against the OTW
Upper value in decision step 312. When OTW Upper value is exceeded, the patient is likely prone and exercising, e.g. by swimming. In such a case, the Target Rate is employed in step 210 of Figure 3 as described below. Moreover, the Tilt Average is not updated in step 312.
F~eturning to step 308, if the Tilt Deviation is within the OTW, then the patient is determirled to be stair climbing in step 314. The Tilt Average signal derived in step 202 is then updated in memory in step 316 by the current Measured Tilt.
Turning to Figure 5, the ,1;~ ", rate calculation step 208 r.~ P~ with the stair climbing decision step 314. If stair climbing was determined in step 310. then the Stair Climbing Rate is compared to the Target Rate in decision step 316. If either "not stair climbing" is determined in step 318 or if the Target Rate exceeds the Stair Climbing Rate in step 320, then the instruction "do not change target rate" is generated in step 322, 2 5 and the Target Rate is employed in step 210. This is also the path that would be followed from step 302 or step 312 if the Tilt Deviation exceeds the OTW Upper value.
In step 320, if the Stair Climbing Rate exceeds the Target Rate, then the Tilt De~ iation is compared to the ITW in step 324. If the Tilt Deviation is within the ITW, then the pacing rate is changed to the Stair Climbing Rate in step 326. If the Tilt 3 0 De~ iation is outside the ITW, then the pacing rate is changed to an il~ JOla;i~JII of the two rates in step 328 through the use of a look up table of interpolation values stored in the memor~v 68 or a r^~ fir~n.

w096/30079 p~ l?`r~ 219~159 i~ x"~ 4 Turning to Figure 6, it depicts one example of the OTW and ITW signal ranges that can be employed to determine a Pacing Rate control signal selected from the Target Rate a bigher (in this example) Stair Climbing Rate, and ;,.' ,.~ ' rates in between these two flat rates. The Tilt Window Deviations in mV vary from the nominal "0" value or l.000 volts as described above. The Target Rate applies outside the OTW Lower and OTW Upper range values of the Tilt Deviation. A higher Stair Climbing Rate applies within the ITW Upper and ITW Lower range values of the Tilt Deviation. The i,lt.~ ; " " I ;~ I r rates prevail in the ramge of Tilt DeYiation values between the ITW Lower and OTW Lower values as well as the ITW Upper and ITW Upper values.
In Figure 6, the Stair Climbing Rate of 120 bpm exceeds the Target Rate of 90 bpm, and therefore controls, if the Tilt Deviation is within the ITW. Tilt Deviations of 20 mV and l l 0 mV, for example, fall outside both the Lower and Upper limits of the OTW
and ITW, and the pacing rate is therefore controlled by the Target Rate of 90 bpm. A Tilt Deviation of 120 mV falls within the ITW, and the pacing rate is controlled by the Stair Climbing Rate (SCR) of 1 2Q bpm. Within the lower portion of the OTW, between 30-60 mV, the i~llr~ pacing rate control signal falls between 90-120 bpm. A formula for t1f-t~-~Tn jnin~ the; "~ I ;,l l r pacing rate (IPR) control signal in this range is:
IPR = (SCR - TR.)(Tilt Devi~fion - OTWL! + TR
2 o (ITWL - OTWL) Similarly, between 80-100 mV, the pacing rate control signal falls bet~een 120-90 bpm. A formula for ~ t~rrninin~ the i" Ill~ r pacing rate (IPR) control signal in this 25range is:
IPR = (SCR - TR)(OTWU - Tilt Devi~tior~) + TR
(OTWU - ITWU) From the above description. it may be seen that the 1i~ m between stair climbing and stair descending or ~halking on a flat surface is sufficient without making a positive ~ that the patient is descending steps or a steep incline. Tbe Target Rate suffices as a stair or incline descending pacing rate. If the Target Rate reflects rapid 35 movement, then it will su~lce for any of the three activities Wo 96/30079 ~ f , ., f~ 21 2 ~ , . 4 Referring back to Figure l, the Target Rate, Stair Climbing Rate or the ihlt~l. ' pacing rate derived in this fashion provide pacing rate control signals derived from the ,l. f, . " ,;. ,,.l ;.-" of the patient body posture and the patient activity level correlated to the ~ ;olOgiC demand on the patient's heart from which physiologic escape intervals 5 rctahlichin~ the physiologic pacing rate are developed by the digital controller/timer circuit 40 and the lll;.lV.Ulll,VUL~l circuit 34 operatirlg as described above. The pacing pulses are generated by the output amplifier circuit 36 at the physiologic pacing rate, and are applied to the patient's heart l 0 through the leads l 2 and 14.
Turning now to Figures 7-9, they depict, in graphical form, the Tilt Deviation 10 /1icfrihllti nnS achieved in l 9 strap-on tests of test subjects derived using the above process during stair climbing, stair descending and normal walking. The three ,li~, ;l,ui;.,,,~
rl~m~ the sensitivity of vhe ,1: ~. ",, .;""i ;"" that can be achieved from the DC
component of the output signal of the DC dc.el~.vl..., L..
The data deri-~ed from a strap-on test of a volunteer subject engaged in walking, 15 restirlg, climbing upstairs, resting, descending stairs and again resting over time is also depicted in Figure 10. The target pacing rate (without .1;~, ,",;"~:,..,.) and the stair climbing rate (with (l;~ nn) generated following the above algorithm from the output signals of the A-P DC acc~l..v.l..,tLI 74 are depicted. As carl be seen, the stair climbing rate increase is ~ JlVlJl- ' Iy effected without illa~)~lvlJl;at~ly increasing the stair 2 o descending pacing rate.
It should be noted that the stair climbing detection may also trigger storage ofepisodic data in Illi-lUCUlll~Ji-l circuit memory for later telemetry out arld analysis by the physician. The r~vlvlff ' of the Stair Climbing Rate and the ITW and OTW
windows can be anal~ zed for the particular patient. The stair climbing d;~LI 1~ of i5 the present invention may be l~lv~l~llmcl offin the event that the patient does not benefit from it.
Variations and mn~lifir~tionc to the present invention may be possible given theabove disclosure. For example, the present invention is not limited to any particular pacing mode, and can function with prior art modes such as DDDR, AAIR, VVIR and 3 o DDIR. It will also be understood that the present invention may be ;. ,~ " ,l~ in dual-chamber rar.-nn ~ PrC cardioverters, ~ ihrillaf~-rc and the like. However, all such variations and m~u1ifi~ationc are intended to be within the scope of the invention claimed by this letters patent.

Claims (10)

1. Apparatus for pacing a patient's heart at a pacing rate dependent on posture,comprising:
means for deriving a body posture tilt signal indicating the degree to which thepatient is leaning forward;
means for defining escape intervals to establish a pacing rate, said escape interval defining means comprising means for altering said escape intervals in response to said posture tilt signal indicating that the patient is leaning forward;
means for generating pacing pulses at the pacing rate; and means for applying the pacing pulses to the patient's heart.

2. The apparatus of Claim 1, further comprising means for detecting patient activity and for deriving a patient activity signal having a signal level dependent on patient activity and wherein said escape interval defining means comprises means for defining escape intervals responsive to both the body posture tilt signal and the patient activity signal.

3. The apparatus of Claim 1 or Claim 2 wherein the means for deriving a body posture tilt signal further comprises means for measuring the forces of acceleration in the anterior-posterior axis of the patient's body recurring in a first predetermined low frequency range and deriving a measured tilt signal therefrom.

4. The apparatus of Claim 2 wherein the means for detecting patient activity further comprises means for measuring the forces of acceleration in the anterior-posterior axis of the patient's body recurring in a second predetermined frequency range higher in frequency than said first predetermined low frequency range and deriving the patient activity signal therefrom.

5. The apparatus of Claim 2 wherein the means for deriving a body posture tilt signal and the means for detecting patient activity further comprise:
means for responding to the force of gravity on the patient's body along a generally anterior-posterior axis of the patient's body and deriving a forward tilt signal therefrom of a polarity and magnitude representative of the degree to which the patient is leaning forward from an upright stance; and means for measuring forces of acceleration in the patient's body due to patient footsteps and deriving a patient activity signal therefrom representative of patient footsteps.

6. The apparatus of Claim 2 wherein the means for defining escape intervals further comprises:
means for providing a target rate control signal for pacing the patient's heart as a function of the patient activity signal;
means for providing a stair climbing rate control signal;
means for defining a tilt window;
means for comparing the body posture tilt signal to the tilt window;
means for employing the stair climbing rate control signal to define escape intervals when the body posture tilt signal is within the tilt window; and means for employing the target rate control signal to define escape intervals when the body posture tilt signal is outside the tilt window.

7. The apparatus of Claim 6 further comprising means for comparing the target rate control signal with the stair climbing rate control signal and employing the greater as the pacing rate control signal.

8. The apparatus of Claim 2 wherein the means for defining escape intervals further comprises:
means for providing a target rate control signal for pacing the patient's heart as a function of the patient activity signal;
means for providing a stair climbing rate control signal;
means for defining inner and outer tilt windows;
means for comparing the body posture tilt signal to the tilt windows;
means for employing the stair climbing rate control signal to define escape intervals when the body posture tilt signal is within the inner tilt window;
means for employing the target rate control signal to define escape intervals when the body posture tilt signal is outside the outer tilt window; and means for deriving an intermediate rate control signal and employing the intermediate rate control signal to define escape intervals when the body posture tilt signal is within the outer tilt window and outside the inner tilt window.

9. The apparatus of Claim 8 further comprising means for comparing the target rate control signal with the stair climbing rate control signal and employing the greater to define escape intervals whether or not the body posture tilt signal is within the inner or outer tilt windows.

10. The apparatus of claim 1 or claim 2 wherein said tilt signal deriving means comprises a DC accellerometer.