WO2007127675A1 - Method and apparatus for verifying a determined cardiac event in a medical device based on detected variation in hemodynamic status - Google Patents

Abstract

A method and apparatus for verifying a determined cardiac event in a medical device based on detected variation in hemodynamic status that includes a plurality of sensors sensing cardiac signals, and a physiologic sensor sensing physiologic signals to generate a plurality of variation index samples corresponding to the sensed signals. A microprocessor detects a cardiac event in response to the sensed cardiac signals, computes a variation index trend associated with a predetermined number of variation index samples of the plurality of variation index samples, determines a rate of change of the computed variation index trend, and confirms the determined cardiac event in response to the computed variation index trend and in response to the determined rate of change.

Description

]

METHOD 4ND APPARATUS FOR VERIFYING A DETERMINED CARDIAC EVENT IN A MEDICAL DEVICE BASED ON DETECTED VARIATION IN

HEMODYNAMIC STATl' S

5 FIELD OF TUE INVENTION

The present invention relates generally to medical devices, and more particularly to a method and apparatus for confirming detection of a cardiac event based on the detection "variations in hemodynamic status using an optical sensor

10 BACKGROU N^rD OF THE l NVEN HON

Implantable medical devices (IMDs) for monitoring a physiological condition or delivering a therapy typically rely on one or more sensors positioned in a patient's blood vessel, heart chamber, oi other portion of the body Examples of such medical devices include heart monitors, pacemakers, implantable cardioverter-defibrillators (KDs),

15 myostimulators, nerve stimulators, drug delivery' devices, subcutaneous defibrillators, and other IMDs where such sensors are desiiable Implantable sensors used in conjunction with an IMD generally prσvide a signal related to a physiological condition from which a patient condition or the need for a therapy can be assessed

Measurement of blood oxygen saturation levels are of interest in determining the

20 metabolic state of the patient Generally, a decrease in blood oxygen saturation is associated with an increase in physical acthϊt\ or may reflect insufficient cardiac output or respiratory activity Thus monitoring blood ox> gen saturation allows an implantable medical device to respond to a decrease in oxygen saturation, for example by pacing the heart at a higher rate An implantable oxygen sensor for use with an implantable medical

25 de\ ice is generally disclosed in commonly assigned U S Pat Ko 6, 198,952 issued to

Miesei, hereby incorporated herein by reference in its entirety Cardiac pacemakers that respond to changes in blood oxygen saturation as measured by an optical sensor are generally disclosed in ϊ ¹ S Pat No 4,202339 issued to Wϊrtzfeld and in L^r S Pat No 4.467.807 issued to Born?in

30 Practical applications for optical hemodynamic sensors, how e\ er, ha\ e been limited because such sensors are highly susceptible to motion, that is, movement by the patient or of the sensoi tends to introduce significant noise onto an output signal of the sensor 2

BRIEF DESCRIPTION OF TOE DRAWINGS

Aspects and features of the present invention will be appreciated as the same becomes better understood by reference to the following detailed description of the embodiments of the invention when considered in connection with the accompanying drawings, wherein: 5 FlG i is a schematic diagram of an exemplary medical device in which the present invention may be usefully practiced;

FIG. 2 is a schematic diagram of an optical hemodynamic sensor according to an embodiment of the present invention;

FlG 3 is a schematic diagram of electronic circuitry included in the device of FIG. 10 1 according to an embodiment of the present invention;

FlG 4 is a flow chart of a method of delivering a therapy in a medical device according to an embodiment of the present invention;

FlG 5 is a graphical representation of identification of an exemplar)- CK variation index trend utilized in a method of delivering a therapy in a medical device according to 15 an embodiment of the present invention;

FIG 6 is a flow chart of a method of delivering a therapy in a medical device according to a.n embodiment of the present invention;

FlG 7 is an exemplary O^ \ ariation index trend identified in a method of delivering a therapy in a medical device according to an embodiment of the present 20 invention;

FIG. S is an exemplary graphical representation of an Oj variation index trend utilized in a method of delivering a therapy in a medical device according to an embodiment of the present invention;

FlG 9 is a (low chart of a method of delivering a therapy in a medical device 25 according to an embodiment of the present invention;

FlG 10 is an exemplar)' graphical representation of an O> variation index trend utilized in a method of delivering a therapy in a medical device according to an embodiment of the present invention,

FIG 1 1 is an exemplar)- graphical representation of generation of a corrected Ch

30 variation index, trend offset utilized in a method of delivering a therapy in a medical device according to a.π embodiment of the present invention; FlG 12 is an exemplary graphical representation of an O> variation index trend utilized in a method of delivering a therapy in a medical device according to an embodiment υf the present invention,

FlG. !3 is a flow chart of a method of delivering a therapy in a medical device 5 according to an embodiment of the present invention,

FlG 14 is an exemplary graphical representation of an <>> variation index trend utilized in a method of delivering a therapy in a medical device according to an embodiment of the present invention; and

FlG i 5 is a flow chart of a method of delivering a therapy in a medical device 10 accordi ng to an embodiment of the present invention .

DE TAILED DESCRIPTION⁷ OF THE LNVENTION

FlG ! is a schematic diagram of an exemplary medical device in which the present invention may be usefully practiced. As illustrated in FlG, 1, the present invention may be

15 utilized in an implantable medical device 14 that includes a housing 15 containing circuitry for operating device 14 that is subcutaneousty implanted in a patient, outside the ribcage of patient 12, anterior to the cardiac notch, for example According to an embodiment of the present invention, housing 15 may be implanted in the pectoral region of the patient 12 Further, device 14 may include a subcutaneous sensing and

20 cardioversion/defibrillation therapy delivery lead 18 coupled to the device 14 that is tunneled subcutaneous!}" into a location adjacent to a portion of a latissimus dorsi muscle of patient 12. Specifically. lead S 8 is tunneled subcutaneous! y from the median implant pocket of device 14 laterally and posterially to the patient^'s back to a location opposite the heart such that the heart 10 is disposed between the device 14 and the distal electrode coil

25 24 and distal sensing electrode 26 of lead 18.

It is understood that while the subcutaneous device 14 is shown positioned through loose connective tissue between the skin and muscle layer of the patient, the term "subcutaneous device^" is intended to include a device that can be positioned in the patient to be implanted using any non-intravenous location of the patient, such as below the

30 muscle layer or within the thoracic cavity, for example.

Further referring to Fisi 1. programmer 20 is shown in teleroetric communication with SubQ ICD 14 by RF communication link 22. Communication link 22 may be any appropriate RF link such as Bluetooth, WiFi. MfCS, or as described in Li S Patent No 4

5,683,432 " Λdapύve Perfonnanee-ϋptinK/ing Coin muni cation S\ stem for Communicating with an Implantable Medical Device^" to Goedeke, et al and incorporated herein by teferencc in its entuetv

Device 14 may be constructed of stainless steel, titanium or ceramic as described 5 in V S Patent \os 4,180,078 "Lead Connector for a Bodv Implantable Stimulator" to

Anderson and S 470,345 'implantable Medical Device w ith Multi-layered Ceramic Enclosure" to Hassler. et al, both incorporated herein by reference in their entireties The electronics chcuitry of device 14 ma> be incorporated on a poly amide flex circuit, printed circuit board (PCB) or ceramic substrate with integrated circuits packaged in Seadless chip

10 carriers and/or chip scale packaging (CSP)

Lead 18, which is inserted within a connector 25 positioned on housing 15 to electrically coupled lead to the circuitry located in housing 15. includes a distal defibrillation coil electrode 24₅ a distal sensing electrode 26, an insulated flexible lead body and a proximal connector pin 27 for connection to housing 15 via connector 25

15 Distal sensing elecUode 2f» is sized appropriately to match the sensing impedance of one or more electrodes 2 S that are positioned along housing 15 to form a housing -based subcutaneous electrode array with electrodes 28 positioned to form orthogonal signal vectors

Dev ice 14 in is an exemplary graphical representation of an Oj variation index

20 trend utilized in a method of delivering a therapy in a medical device according to an embodiment of the present invention includes an optical sensor 17 positioned along the outer surface of housing S 5, which is utilized to generate an Cb \asiation index trend for use in generating a secondary confirmation of the detection of a cardiac event by the primary detection algorithm, as described in detail below Flectrodes 28 and optica!

25 sensor 17 are welded into place on the outer surface of the housing 15 and are connected via wires (not shown) to elecUonic circuitry (described herein belou) located inside housing 15 hlectrodes 28 may be constructed of flat plates, or alternatively, spiral electrodes as described in U S Patent No 6,512,940 "Subcutaneous Spiral Electrode for Sensing Electrical Signals of the \ ϊearf^" to Brabec, et a! and mounted in a non-conductive

30 surround shroud as described in U S Patent Nos 6,522,015 "Sunound Shroud Connector and Electrode Housings for a Subcutaneous Electrode Array and ϊ.eadlesa HCGs"^' to Ceballos, et al and 6,622,046 "Subcutaneous Sensing Feedthrough/Eleetrode Assembly" to Fraley, et al, all incorporated herein by reference in their entireties The electronic circuitry employed in device 14 can take any of the known forms that detect a tachyarrhythmia from the sensed ECG and provide cardioversion defibrillation shocks as well as post-shock pacing as needed while the heart recovers Λn exemplars simplified block diagram of such circuit adapted to function 5 employing the first and second cardioversion-defibrillation electrodes as well as the ECG sensing and pacing electrodes described herein belcm is set forth in KIG 3 It will be understood that the simplified block diagram does not show all of the conventional components and circuitry of such devices including digital clocks and clock lines, low voltage power supply and supply lines for powering the circuits and providing pacing

10 pulses or telemetry circuits for telemetry transmissions between the device 14 and external programmer 20

Optical hemodynamic sensor 17 is preferably a multiple waveform oximeter, such as a pulse oximeter or a niixed-\enous ox\gen sensor, for example Pulse oximeters are well known sensors commonly used with various medical devices, both implantable and

15 external Foi example some applications of optical oximeters are disclosed in I l S patent

Nos 4.750.495, 5, 176,137, 6. i 44.866. 6J 98,952, or 6.944.488. each of which is assigned to Medtronic, lnc

Generally, optical oximeters include a light source for emitting light through a blood perfused tissue of patient P and a light detector for generating a signal representative

20 of an intensity of light transmitted through the blood perfused tissue to the light detector

In other embodiments, the mixed-venous ox> gen sensor may be placed in the blood stream itself Che light passed through the tissue or bloodstream is commonh selected to he of two or more wavelengths, and most commonl}, the light is selected to fall in the red part of the visible light spectrum and the infrared (IR) portion of the light spectrum The light

25 transmitted through the blood perfused tissue or bloodstream and received by the light detector is general!) reprcsentati\ e of hemodynamic function

FlG 2 is a schematic diagram of an optical hemodynamic sensor according to an embodiment of the present invention 4.s illustrated in FJG 2 optical sensor 17 includes a red light emitting diode UJKD) 21. an infrared (ΪR) LHD 23, a photodiode 30. and an

30 optical barrier 32. all of which are positioned within a sensor housing 34 having a lens 3fr ϊn the embodiment shown in FlG 2, LCDs 21 and 23 and phoiodiode 30 are each mounted on a substrate 37, or a bottom surface of housing 34 As indicated by arrows 38, red and Hl LHDs 21 and 23 are configured to emit light through lens 36 of housing 34, while, as 6 indicated b) arrows 39_} photodiode 30 is configuied to detect light receiv ed through lens 36 Optical barrier 32 is positioned to Mock direct transmission of light from I PDs 21 and 23 to photodiode 30

Optical hemoch naittie sensor 17 preferably is subcutaneous Iy or submuseularSy 5 implanted within patient P such that lens 36 is oriented toward a blood perfused tissue of patient P

Red LED 2 S preferably emits light in the red portion of the visible light spectrum, while IR. (J-¹I) 23 preferably emits f R. light in the IR portion of the light spectrum In alternate embodiments, optical hemodynamic sensor 17 may include any two or more light

10 sources for producing at least two different wavelengths of light Photodiode 30 preferably recehes light transmitted by LMXs 21 and 23. with an intensity of the signal received by photodiode 30 being indicative of bernoch namtc function For instance, oxj gen saturation of the Mood can be derived from an output of photodiode 30. as \s\\\ be described below, and used to provide a secondary confirmation of a detected event by the

15 device 14 according to the piesent invention

FIG 3 is a schematic diagram of electronic circuitry included in the device of FlG 1 aceoidmg to an embodiment of the present invention As illustrated in FIG 3, device 14 includes both a low voltage battery 153 and a high voltage batten i 12. for example, positioned within the hermetically sealed housing 15 of the device 14 Low voltage

20 battery 153 is coupled to a power supply (not shown) that supplies power to the de\ ice circuitry and the pacing output capacitors to supply pacing energy in a manner well known in the art The low \ oltage batten 153 can include one or more com entional LiCh _v LiMnOj or Lib ceils, while the high voltage batten 1 12 can include one or more conventional IiSVO or Li-MnO; cells It is understood that although the exemplary

25 embodiment of FIG 3 includes both low and high power theraps . the present invention may be employed in a device that provides only one therapy , such as a high powei defibrillation therapy, for example

Dc\ ice 14 functions are conti oiled by means of software, firmware and hardware that cooperative^ monitor the IiCG. determine when a cardioversion-defibrillation shock

30 oi pacing is nccessarv, and deliver prescribed cardioversion-defibrillation and pacing therapies FlG 3 incorpoiates eύcuitry set forth in common!) assigned U S Patent Nos 5 J 63.427 "Apparatus for Delivering Single and Multiple Cardioversion and Defibrillation Pulses^" to Keimel and 5,188, 105 "Apparatus and Method foi Treating a Tachy arrhythmia" 7 to Ketmel for selectively delivering single phase, simultaneous bipbasic and sequential bi phasic cardioversion-defibrillation shocks, incorporated herein by reference in their entireties

In FKx 3, sense amp 190 in conjunction with pacer'device timing circuit 178 5 processes the far field ECG sense signal that is developed across a particular ECG sense

\ ector defined by a selected pair of the subcutaneous eieeuodes 28 oi, optional)} , a s irtuai signal if selected The selection of the sensing electrode pair is made through the switch matrix /M UX !91 in a manner to provide the most reliable sensing of the KGM signal of interest, which would be the R. wave for patients who are believed to be at risk of

10 ventricular fibrillation leading to sudden death The far field ECG signals are passed through the switch niatrix/MUX 1 *51 to the input of the sense amplifier 190 that, in conjunction with pacer/device timing circuit 178. evaluates the sensed FGM Bradycardia, or asv stole, is typically determined b\ an escape interval timer within the pacer timing circuit 178 and/ or the control circuit 144 Pace Trigger signals are applied to

15 the pacing pulse generator 192 generating pacing stimuiaiion when the interval between successive R- waves exceeds the escape inteπ. al Brad> cardia pacing is often temporarily provided to maintain cardiac output aftei delis ery of a cardiox ersion-defibrillation shock that may cause the heart to slowly beat as it recoxers back to normal function Sensing subcutaneous far field signals in the presence of noise may be aided by the use of

20 appropriate denial and extensible accommodation periods as described in Ii S Patent No

6,236,882 "Noise Rejection for Monitoring ECGs^" to Lee, et al and incorporated herein by reference in its^* entirety

Detection of a malignant tachyarrhythmia is determined in the control circuit 1 Φi. for example, as a function of the intervals between R-wave sense event signals that are

25 output from the pacer/device timing 178 and sense amplifier circuit i*X) to the timing and control circuit 144

Supplemental sensors such as tissue color, tissue oxygenation, respiration, patient acti\ ity and the like may be used to contribute to the decision to apply or withhold a defibrillation therapj as described generally in U S Patent No 5,464.434 "Medical

30 Interventional Device Responsive to Sudden Hemodynamic Change" to 4.It and incorporated herein bj reference in its entirety Jn particular, the present invention includes optical sensor 17 to provide a secondary confirmation of a detected tachyarrhythmia event detected b\ the device 14 bv determinina; whether the heart is S hemodynamically unstable m response to a tachs cardia event being identified by the dev ice 15 in response to ll-wa\e sense intervals determined in the prinian detection algorithm, described below in detail Sensor processing unit 194 provides sensor data to microprocessor 142 via data bus 146 In addition to optica! sensor ! ?. an acthϊt\ sensor 5 may also be utilized so that patient activity and/or posture may also be determined by the apparatus and method as described in U S Patent No 5,593,43 i "Medical Service Employing Multiple DC Accelerometers for Patient Λcthitv and Posture Sensing and Method^" to Sheldon and incorporated herein by reference in its entirety Similar!) , patient respiration may be determined by the apparatus and method as described in U S Patent

10 No 4,567,892 "Implantable Cardiac Pacemaker" to Plicchi. et al and incorporated herein by reference in its entirety Λs mentioned above, according to the present invention. optical sensor ! 7 ma\ be located on the housing \ 5 of device 54. or may be located on the lead 18 to enable the sensing of contacting or near-contacting tissue oxygenation Certain steps m the performance of the detection algorithm criteria are

15 eooperam el y performed in microcomputer 142 including micropiocessor, RAM and

ROΛf, associated circuitry, and stored detection criteria that may be programmed into RAM via a telemetry interface (not shown) con\ enttona! in the art Data and commands are exchanged between microcomputer 142 and timing and control circuit 144, pacer timing/amplifier circuit 178, and high v oltage output circuit 140 via a bi-directional

20 data/control bus !46 The pacer timing/amplifier circuit 178 and the control circuit !44 are clocked at a slow clock rate The microcomputer 142 is normally asleep, but is awakened and operated bv a fast clock by interrupts developed bv each R-wave sense event, on receipt of a downlink telemetry programming instruction or upon delhery of cardiac pacing pulses to perfoim aπ> necessary mathematical calculations, to perform

25 tachycardia and fibrillation detection procedures, and to update the time intervals monitored and controlled by the timers in pacer/device timing circuitry 178

The algorithms and functions of the microcomputer 142 and control circuit 144 employed and pci formed in detection of tachyarrhythmias arc set forth, for example, in commonly assigned V S Patent Nos 5.354.316 "Method and Apparatus for Detection and

30 Treatment of Tach\ caidia and Fibrillation"' to Keimcl. 5,^45.186 "Prioritized Rule Rased

Method and Appaiatυs for Diagnosis and Treatment of Arrhythmias"^" to Olson et al, 5.855.593 "Prioritized Rule Based Method and Apparatus for Diagnosis and Treatment of Arihs thmias^"1 to Okon, et a! and 5, 193,535 "Method and Λppaiatus for Discnmination of Ventricular Tachycardia from Ventrieulai Fibrillation and Treatment Fhereof¹ to Bardy, et al (all incorporated herein b> reference in their entireties) Particular algorithms for detection of ventricular fibrillation and malignant v entricular tachycardias can be selected from among the comprehensive algorithms for distinguishing atrial and ventricular 5 tachyarrhythmias from one another and from high rate sinus rrr> thms that are set forth in the \^"S1o - 186, \593 and '5^3 patents

The detection algorithms are high!\ sensitive and specific for the presence or absence of life threatening ventricular anhythnuas, e g , \ eπtπcular tach>caidia (VT) and ventricular fibrillation (VF) When a malignant tachycardia is detected, high \ oltage

10 capacitors 156, 158, 160, and 162 are charged to a pre-programmed voltage level by a high-voltase charging circuit 164 It is general Iv considered inefficient to maintain a constant charge on the high voltage output capacitors 1 %, 1 58. 1 frO, 162 Instead. charging is initiated when control circuit 144 issues a high \oltaae cbarae command HVCHG delivered on line 145 to high voltage charge circuit it>4 and charging is

15 controlled by means of bi-directional control/data bus 166 and a feedback signal VCAP from the HV output circuit 140 High \ oltage output capacitors 156, 158, 160 and 162 ma\ be of film, aluminum electrolytic or wet tantalum construction

I he negative terminal of high voltage battery 1 12 is directh coupled to system ground Switch circuit 1 14 is normally open so that the positive terminal of high voltage

20 battery 1 12 is disconnected from the posith e power input of the high \ oltage charge circuit 164 The high voltage charge command HVCHG is also conducted via conductor 149 to the control input of switch circuit 1 14, and switch circuit 1 14 closes in response to connect positive high voltage battery voltage EXI B- to the positive power input of high voltage chaige ciicuit 164 Su itch circuit 1 14 may be, for example, a field effect

25 transistor (FET) with its source-to-drain path interrupting the EXT B+ conductor 118 and its gate receiving the HVCHG signal on conductor 145 High voltage charge circuit 164 is thereby rendered readv to beuin charging the hi ah. voltage output capacitors 156. 158, 160. and 162 with charging cur tent from high voltage batten* 1 12

High voltage output capacitors 156. 158, 160. and 162 maj be charged to ver\ high

30 voltages, e g , 700-3150V, to be discharged thtough the body and heart between the electrode pah of subcutaneous cardioversion-defibrillation electrodes 1 S3 and 123 The details of the voltage charging circuitry are also not deemed to be critical with regard to practicing the present invention, one high voltage charging circuit belies ed to be suitable io for the purposes of the present invention is disclosed High voltage capacitors 156, 1.58, 160 and 162 are charged by high voltage charge circuit 164 and a high, frequency, high- voltage transformer 168 as described in detail in commonly assigned U.S. Patent No. 4,548,209 "Energy Converter for Implantable Cardioverter" to Wieiders, et al. Proper 5 charging polarities are maintained by diodes 170, 172, 174 and 176 interconnecting the output windings of high-voltage transformer 1.68 and the capacitors 156, 158, 160, and 162, As noted above, the state of capacitor charge is monitored by circuitry within the high voltage output circuit 140 that provides a VCAP, feedback signai indicative of the voltage to the timing and control circuit 144. Timing and control circuit 144 terminates

10 the high voltage charge command HVCHG when the VCAP signal matches the programmed capacitor output voltage, i.e., the cardioversion-defibrillation peak shock voltage.

Control circuit 144 then develops first and second control signals NPULSE 1 and NPULSE 2, respectively, that are applied to the high voltage output circuit 140 for

15 triggering the delivery of caxdi overling or defibrilialing shocks. In particular, the

NPULSE i signal tri ggers discharge of the first capacitor bank, comprising capacitors 156 and 158. The NPULSE 2 signal triggers discharge of the first capacitor bank and a second capacitor bank, comprising capacitors 160 and 162. It is possible to select between a plurality of output pulse regimes simply by modifying the number and time order of

20 assertion of the NPULSE 1 and NPULSE 2 signals. The NPULSE 1 signals and NPULSE

2 signals may be provided sequentially, simultaneously or individually. In this way, control circuitry 144 serves to control operation of the high voltage output stage 140, which delivers high energy cardioversion-defibrillation shocks between the pair of the cardioversion-defibrillation electrodes 1 13 and 1.23 coupled to the HV-I and COMMON

25 output as shown in FIG. 3.

Thus, device 14 monitors the patient's cardiac status and initiates the delivery of a cardioversion-defibrination shock through the cardioversion-defibrillation electrodes 1 13 and 123 in response to detection of a tachyarrhythmia requiring cardioversion- defibrillation. The high HVCHCϊ signal causes the high voltage battery 1 12 to be

30 connected through the switch circuit 1 14 with the high voltage charge circuit 164 and the charging of output capacitors 156, 1.58, 160, and 162 to commence Charging continues until the programmed charge voltage is reflected by the VCAP signal, at which point control and timing circuit 144 sets the HVCHG signal low terminating charging and π opening witch circuit 1 14 T) pica!!) , the charging cycie takes on!) fifteen to twenty seconds, and occurs \ery infrequently lhe device !4 can be programmed to attempt to deliv cr cardioΛ ersiαn shocks to the heart in the manners described above in timed synchrony with a detected R-wave or can be programmed or fabricated to del her 5 defibrillation shocks to the heart in the manners described above without attempting to synchronize the delivery to a detected R-v,a\e episode data related to the detection of the tachyarrhythmia and delivery of the cardioversion-defibrillation shock ma> be stored in RAM for uplink telemetry Uansmission to an externa! programmer as is well known in the art to facilitate in diagnosis of the patient's cardiac state A patient receiving the device i 4

10 on a prophylactic basis would be instructed to report each such episode to the attending physician for further evaluation of the patient's condition and assessment for the need for implantation oi^'a mote sophisticated implantable cardio-dcfibriliator device (ICD) In other embodiments, no storage of episode data will take place

FlG 4 is a flow chart of a method of delivering a therapy in a medical device

15 according to an embodiment of the present invention As illustrated in FlG 4 once control circuit 144 determines the presence of a malignant cardiac event using the primary detection algorithm described above. Block 400, the piesent in\ ention generates a secondary confirmation of the event detected h\ the primary detection algorithm, blocks 402-410 In particular, once the primary detection algorithm is satisfied, the ptesent

20 invention uti Iszes the input generated from optical sensor 17 at multiple wa\ elengths to identify an O; variation index trend, block 402, as described below in detail According to the present invention, the O; \ ariation index trend is a measure of the change in tissue ox\genation and corresponds to the relationship between changes in both the volume of blood at the sensor site and the concentration of oxygenated hemoglobin (i ϊfa O;)

25 I'sing the results of the generated O; variation index trend, a determination is made as to whether the detected event is associated with noise, resulting ftom patient motion, for example. Block 404 If it is determined that the detected event is associated with noise, delivery of therapy is withheld, or control of the device is reverted back to the primary detection algorithm, Block -40(> However, if it is determined that the detected event is not

30 associated with noise, a determination is made as to whether the detected event is associated with an unstable rhythm. Block 408, such as \ eπtrieular tachycardia or ventricular fibrillation, for example If the detected event is not determined to be associated with an unstable i"h)thm, the result of the primary detection scheme is not confirmed and delivery of therapy is withheld or contfol of the device is r everted back to the primary detection algorithm. Block 406 If the detected event is determined to be associated with an unstable rh\ thin. 5 the primary detection of the malignant event is confirmed. Block 4 J 0. and therapy is delivered

MG 5 is a graphical representation of identification of an exemplary Ch v ariation index U end utilized in a method of delivering a therapy in a medical device according to an embodiment of the present invention As illustrated in FlG 5, once control circuit 144 10 determines the presence of a cardiac event, such as ventricular tachycardia or ventricular fibrillation, for example, using the primary detection algorithm described above, the presence of the cardiac event is confirmed by determining changes in blood oxygenation of the patient The changes in Mood oxygenation are determined using an O; \ ariation index trend 200 that is identified based on the intensity readings associated with the 15 intensity of the red light emitted bj red LCD 21 and the infrared light emitted by infrared

LED 23 that is received at photodiode 30

In particular, in oidei to identify the ϋ; \ ariation index trend 200, both a red light baseline intensity I₁, and an infrared light baseline intensity i\ is identified from sample outputs received at a predetermined sample rate over a sampling time interval For 20 example, according to an embodiment of the present invention, sample outputs are received at photodiode 30 from red LKD 21 and infrared LFD 23 at a sampling rate of three samples per second over a two second sampling time interval Baseline intensity /_t and baseline intensity /^* are then determined from the sample outputs from red LFD 21 and infrared LED 23, respcetiv cly For example, according to an embodiment of the 25 present in\ention, baseline intensity /,, and baseline intensity /^* are determined, respectively . by setting baseline intensity /■> equal to the average of the sample outputs from red LHD 21 over a predetermined time period and setting baseline intensity /^* equal to the average of the sample outputs from infrared Lliϋ 23 over the predetermined time period Once the ied and mfraied baseline intensities /_t, and /," base been determined a variation index is determined for each subsequently receiv ed two-wavelength sample output b> sensor i 7 using an ovygen variation index equation

liquation 1 Variation index JL- L i. Λ

where / is the intensity of red light from red LBD 21 incident on photodetector 30 for a given sample and ι^" in the intensity of infrared light from infrared LED 21 incident on

10 photodetector W for the same sample In this way, the variation index for each two- wavelength sample output is the difference between the proportion of the red and the infrared intensity signals with respect to their corresponding baseline intensities Using the exemplary sampling rate of 3 H/, three variation indexes are generated each second and are used to determine the C> variation index, trend 200

15 Ii is understood thai other relationships between the proportions of the red and infrared intensities to their corresponding baseline intensities ]_ and >^* For example, i f the propoi tiυn of the red intensity signal to the ' ^1<S baseline red intensity ~~ is referred to as the normalized red intensit\ and the proportion of the infrared intensity signal to the baseline infrared intensity ^~τ

20 is referred to as the normalized infrared intensity. Equation I rna\ alternatively be a ratio of the normalized red and infrared intensities, or ma\ be a difference between the unequally weighted red and infrared normalized intensities

FIG 5 includes three oxygen variation index trends identified using the oxygen variation index equation, Equation 1 The fust is an exemplary O₃ variation index tiend resulting

25 during norma! sinus rhythm 202. the second is an exemplary O> variation index trend resulting during noise 204. such as patient motion for example, and the third is an exemplar} O- variation index trend resulting during a malignant cardiac event 206. such as \ entriciiSar fibrillation for example 4.S can be seen in FlG 5, using the O; variation index trend 200 identified according to the present invention, the C); variation index trend

30 resulting during noise 204 tends to exhibit a v ariability that is significantly greater than the variability exhibited bv the Oj vacation index tiend resulting during normal sinus rhythm 202, while the O> variation index trend resulting during the malignant cardiac event 206 Ϊ4 tends to exhibit a variability that is generally equivalent to oτ less than the variability exhibited by the O; \ariaiion index trend resulting during normal sinus rhythm 202 Λs described below, the present invention utilizes these variation features and others to perform a secondar\ confirmation of a cardiac event determined by a prirnar\ detection 5 scheme to be a malignant cardiac event One such criterion involves crossing of a predetermined baseline 210 bj a patametei, described in detail below

MG 6 is a flow chart of a method of delivering a therapy in a medical device according to an embodiment of the present im ention FlG 7 is an exemplary O; variation index trend identified in a method of delivering a theraps m a medical device according to

10 an embodiment of the present invention As illustrated in FIGS 5 and 7, once the primary detection algorithm is satisfied. Yes in Block 500. the present imention begins with the initialization of several parameters, such as a corrected O; index trend. Block 501 , which is intended to represent a measure of the change in tissue oxygenation and will be utilized below in reference to Block 508

15 Once the initialization of parameters is complete, i e . the corrected O; index, trend is set equal to zero, the present invention begins computing O; \ariation indexes 600 using the optical sample inputs fτora optical sensor 17 at multiple wavelengths and liquation 1. Block 502 Optical sample inputs from sensor 17 are collected and the corresponding O; \ ariatiυn indexes oOO are computed using Equation 1 for a predetermined sample

20 collection period The predetermined sample collection period may be set as a predetermined period of time, such as 5 seconds, for example, or may be set as a predetermined number of samples, such as 15, for example in the an exemplary trend according to the present invention, the sample rate of 3 Hz is utilized and the sample collection period is set as five seconds, for example, so that 15 O; variation indexes 600

25 are determined over each sample collection period

At the end of the initial sample collection period, an O; variation index trend 602 is identified for the sample collection period. Block 503. and a measure of the deviation of each of the acquired O; variation indexes 600 occurring during the current sample collection period from the O2 variation index trend o02 is determined, Block 504

30 According to an embodiment of the present invention, the O; variation index trend 602 is identified in Block 503 bj performing a least square linear fit of the acquired Oj variation indexes 600 during the sample collection period, i e , the first through the 15^th O; \ ariation indexes u^u ~ 0^1:%, so that the resulting trend has a start point 604 where the first O2 15

\ ariation uidex O^ j > is projected onto the C); vacation index trend (>02 and an endpomt 606 where the last O; \ ariation index O;₍i^ is projected onto the O; variation index trend 602, i c , the 15^th O; variation index oOO, and the measure of deviation of the curt em samples from the trend Block 504, is performed b\ determining the mean square 5 dc\ iation of the O; \ ariation indexes 600 in the current window of Oz vanation indexes from the O; \ ariation index trend 602

According to an embodiment of the present invention, the 02 variation index trend 602 ma> be obtained, for example, by an alternative tillering technique and the measure of the deviation of the indexes 600 from the On variation index trend t>02 may he determined

10 as the mean square of the indexes 600 from the filtered index trend

In order to perform the secondary determination of whether the detected event is associated with noise (Block 404 oi^'FIG 4). once the deviation of the associated O? variation indexes 600 from the current O; variation index trend is determined. Block 504, a determination is then made as to whether the deviation is less than a predetermined

15 deviation ilueshoid. Block 505

If the deviation of the C> variation indexes 600 for the current window of Oz variation indexes 600 is not less than or is equal to the deviation threshold, No in Block 505, indicating a likelihood that the determined cardiac event may be the result of noise, a determination is made as to whether a predetermined episode verification time period has

20 expired, Block 50? If the episode verification time period has not expired, the process returns to Block 502 so that the deviation of O; variation indexes from the O? variation index trend, Blocks 503 and 504, is determined for the next window of O; variation indexes 600

In particular, once the deviation of the O; variation indexes 600 for the current

25 window of O^ variation indexes is determined to be greater than the deviation threshold and the episode \ erification time period. No in Block 507, i e , 30 seconds for example, has not expired, an O^ \ ariation index trend 608 is determined for the next window of Oz \ ariation indexes, i c , the window including the 2^1K through the 16'¹ O; variation indexes O;o - Oj^jbv !>o that the resulting trend has a start point 610 w here the second On \ ariation

30 index O^i) is projected onto the Oi variation index trend 60S and an endpoint 612 where the last O; variation index O^im is projected onto the {>> variation index trend 608 The dev iation of the O; v ariation indexes 600 for that window of O: variation indexes O₂^ - <);_{i_b) from the O; variation index trend 608 ts determined and compared to the deviation Ϊ6 threshold, Biøck 505 hi this way, the process of the present invention continues to compute the O; variation index trend over a moving group of consecutive computed Cb variation indexes including some of the most recently acquired samples, with the size of the group of samples being consistent with the sample collection period. 5 The process continues for the next window of Oz variation indexes 500, and if the deviation of generated Cb sanation indexes 600 from the associated trends for the subsequent sample collections periods continues to be greater than the deviation threshold and therefore the episode verification time period has expired, the secondary confirmation process determines that the cardiac event is most likely related to noise, and therapy is

10 withheld or control of the device is reverted back to the primary detection algorithm.

Block 506.

The episode verification time period may be set at any desired value, so that the determination of whether the cardiac event is noise related may be made for a predetermined number of iterations or over a predetermined time period, depending on the

15 chosen values for the sample collection period and the episode verification time period.

For example, according to an embodiment of the present invention, the episode verification time period is set at 30 seconds and the sample col lection period is set at 5 seconds In this example, if the optical sensor signals are sampled a 3 Hz, ninety O; variation indexes are computed over the episode verification time period, with the

20 associated (h variation index trend and deviation being computed every- 5 seconds over a moving group of 15 consecutive samples. Since the computation of the Cb variation index trend begins once 15 variation indexes are computed, the determination of whether a detected cardiac event is noise related, and if not, whether it is hemodynamic stability, Blocks 509 and 5 i 1. is made for seventy-six iterations, over each 30 second episode

25 verification time period.

If the deviation of the O; variation indexes for a given window of O? variation indexes is determined to be less than the deviation threshold. Yes in Block 505. indicating a likelihood that the determined cardiac event is not the result of noise, the corrected O> index trend is incremented by the determined deviation of the current Cb variation indexes,

30 Block 5OS, as will be described in detail below A determination is then made as to whether the current generated Oj variation index trend 602 is less than a predetermined baseline value 210, Block 509. 1 7

According to the present invention, the baseline value 210 associated with Block 509 corresponds to a desired delation associated with the relationship between the proportion of the intensity of red light / from LED 21 to the baseline intensity t₀ fur LED

21, or — of Equation L and the proportion of the intensity of infrared Jmht / from LED

5 23 to the baseline intensity /! for LED 23, or — of Equation 1 For example, according to

>< an embodiment of the present invention, the baseline value 210 is set as -0 02, corresponding to the proportion of the intensity of infrared light /^"" from HID 23 to the baseline intensity /_t ^v for LHD 23 being greater than the proportion of the intensity of red light / from LKD 21 to the baseline intensity /₍, for LPD 21

10 if the current O- v ariation index trend is not determined to be less than the predetermined baseline value 210. No in Block SCO, indicating that while the determined cardiac event is not likely the result of noise, there is a likelihood that the determined cardiac event may be not be associated with an unstable rhuhm, the determination is made as to whether the episode verification time period has expired, Block 507 If the episode

15 verification time period has not expired, the process returns to Block 502 so that the dev iation of O? variation indexes 600 is determined for the novt window of Oj variation indexes, block 503

If the current O2 variation index trend is determined to be less than the predetermined baseline value 210, indicating that the determined cardiac event is not

20 likely the result of noise, and there is a likelihood that the determined cardiac event may be associated with an unstable rhythm, a determination is made as to whether the generated Oi variation index trend is sustained, i e , remains less than the predetermined baseline value 210 for a predetermined time period, such as 3 seconds for example. Block 511

25 If the generated O? variation index trend is not sustained, i e , not less than the predetermined baseline value 210 for the predetermined time period, the determination is made as to whether the episode verification time period has expired. Block 507 If the generated O; variation index, trend is not sustained and the episode v education time period has not expired, the deviation of the generated Cb variation indexes over the next window Ϊ 8 of C>₂ variation indexes is determined. Blocks 503 and 504, and the noise determination is repealed

If the generated O; variation index trend is less than me predetermined baseline value 210 for the predetermined time period, i e . the generated (> variation index trend is 5 sustained, the secondary confirmation process confirms the identification of the malignant cardtac event, and therapy is delivered. Block 512

According to an embodiment of the present invention, the deviation threshold of Block 505 is determined, for example, by periodically computing Oa variation indexes using Equation 1 and generating a corresponding oxygen variation index trend during a

10 known period of motion-free normal sinus rhythm, such as while the patient is sleeping

Hie deviation of the oxygen variation indexes generated during normal sinus rhythm from the corresponding oxygen \ ariation index trend generated foi the sample collection period is then determined using the same process utilized in Block 503. such as the mean square de\ iation, for example Other methods for determining the deviation may be utilized

15 rather than the mean square deviations, such as the mean of absolute values of deviations, for example The deviation threshold utilized for Block 505 is then updated h\ being set equa! to the deviation of the oxygen \ ariation indexes from the trends generated during motion-free normal sinus rh\ thm, or to a multiple or a fraction of the deviation

FlG 8 is an exemplar} graphical ^presentation of an O> variation index trend

20 utilized in a method of delivering a therapy in a medical de\ ice according to an embodiment of the present invention According to an embodiment of the present i mention, during the incrementing of the corrected OJ variation index trend. Block 508, the endpoints of the current determined O; \ ariation index trends are used to calculate a current trend value AOa_1n. associated with the cuirent sample collection period IΌI

25 example, as illustrated in FlG 8, once the O; variation index trend 800 has been determined, and the first and the last O; \ ariation index samples of the 1 S O; variation index samples associated with the window of Oz \ ariation indexes 600 are projected onto the O; variation index trend 800 to determine a first index trend tO^, 1 _> and a last index trend tiha^ associated with the O- variation index trend 800. as described alxne. the

30 change in index trend ^O^^₎ for the O; variation index trend SOO is then determined as the difference between me first O> variation index trend tO^ j> and ihe last C); \ ariauon index trend tθ2_tj ^, i e , tθ2, j ^ - tO^ t ) The corrected O; \ ariation index trend cO^ is then set equal to the determined change in the index trend ΛO;,^₎ A next O; variation index trend 802 is then determined for the subsequent window of t>2 v ariation index samples, i e , the next O; v ariation index sample (Vt, u , and the previous n-1 of the Oj variation index samples, O;,;, through Oαu^t The first and the last O; variation index samples O;^ and ⁽.hti i are then projected onto tine O; variation index 5 trend 802 to determine a first index trend tθ;^j and a last index trend tO^j,,, associated with the O; \ aria.ion tndex trend 802, and the change in the index trend ΛOja_"! for the (h \ariation index trend 802 is then determined as the difference betvseen the first index trend tθ;,2> and the last index trend tO;^, i e , tO;(|_W - tO^,;,

The corrected O^ variation index trend eO^ is then incremented in Block 508 b\

10 being set equal to the sum of the previous corrected O₂ variation index trend and the product of the inverse of the number of samples in the window of O^ variation indexes and the determined change in the Oz variation index trend \O;^_<,) foi the O? variation index trend 8(KJ for the current window of O^ \ ariation index samples I his process is then repeated so that during noise free periods identified in Block 505, the corrected O;

15 variation index trend cθ?(_{i h} is incremented for each window of O; variation index samples b> being set equal to the sum of the prev ious corrected O; variation index trend cCWn and the product of the inverse of the number of samples n in the sample collection period and the determined change in the Oi \ ariation index trend ΔO^,,for the O; variation index trend associated with the current window of O; variation index samples, indicated by the

20 fol I ovv i ng equati on

cθ;_(ι> ~ eθ;_h ,. i j + ! /n(ΔOj_u ,) Equation 2

25 According to another embodiment of the present invention, once the first sample collection period, such as 0-5 seconds for example, has expύed and the corresponding CK variation index trend 800 has been determined, 15 O; variation index trend values tO^i , through tθ>(_tr\ along the O; variation index trend 800 a?e identified by projecting the location of each of the 15 Ch \ ariation index samples for the sample collection period onto

30 the O; variation index trend 800 The change in the O; variation index trend \Q₂, _J M for the O> variation index trend 800 is then determined as the difference between the first index trend tCb t, and the last index trend tO^j^, i e , tO^ I M - tθ;^j 20

An O; variation index trend 802 is then determined for the next window of O? variation index samples, and corresponding trend values tO;_f2-κΛ along the O; v ariation index trend 802 are determined by projecting the location of each of the Cb variation index samples onto the On variation index trend 802 as described above. The change in the CK 5 variation index trend AQ₂, ^₁ for the CK variation index trend 802 is then determined as the difference between the first index trend tCK^ and the last index trend tO^, \&_h i e.. iOja^i - tθ2^). The corrected O; variation index trend is then incremented by being set equal to the sum of the change in the current Oa variation index trend AO^ t_<v> and the product of the inverse of the number of samples n in window of O? variation indexes and the change in

10 the O; variation index trend AOio?) determined for the previous sample collection period, indicated by the equation CO^K.. ~ l/n{Λ0;₍i;>) ^■*^• ΛO^tt,,.

An O; variation index trend 804 is then determined foi the next window ofO? variation index samples, and corresponding trend values tOi^-m along the CK variation index trend S04 are determined by projecting the location of each of the O^ variation index

15 samples associated with the window onto the O? variation index trend 804 The change in the O^ \ aήation index trend ΔCbrt for the O^ variation index trend 804 is then determined as the difference between the first Oj variation index trend tihø₎ and the last O; variation index trend tθ;α> i e , tO^p, - tO^. The corrected O^ variation index trend c.0_{2( t ?}^ is then incremented by being set equal to the sum of the product of the inverse of the number

20 of samples n in the sample collection period and the change in the O; variation index trend

ΛOJ. LM determined for the first O; variation index trend 800, the product of the inverse of the number of samples n in the sample collection period and the change in the CK variation index trend AO^ jt_» determined for the previous O; variation index trend 802 and the change in the O; variation index trend M>₂, r, for the current O; variation index trend 804,

25 indicated by the equation: cθ;u7, = IZn(AO^Hx) -- i Zn(AO^Uu) ^ AO^i?).

Once the corrected C); variation index trend for three noise free sample collections ha\e been determined, an initialization period for the corrected (> variation index trend incrementation in Block 508 is completed, and the corrected O₂ variation index trend cO; is updated for subsequent windows of O; variation index samples identified by the

30 determined value of the last O; variation index trend corresponding to the endpoint of the most recent determined O> variation index trend using the following equation

cO_2u,_α) ^::: cθ_2(!,_n.;, -i- l/n {ΛO_Λ,.,,.n) ^■(^■ AC)^₁.,,, Equation 3 where i is the last O; \ ariation index, trend corresponding to the endpoint of the most ϊecent determined O; variation index trend, cO^i _n-:n is the corrected 0> \ ariation index trend associated with the window of O; \ ariation index samples occurring two sample 5 collection periods prior to the current window of O^ variation index samples, i/n is the inverse of the numbei of samples n in the sample collection period, (Λθj_{u n}.i,) is the corrected Q2 variation index trend associated with the window of O; variation index samples occurring one sample collection peπod prior to the euπent window of O; variation index samples, and AO;,, n,h the change in the current O; variation index trend

10 In both embodiments of the present in\ cnti on associ ated \\ ith Equati ons 2 and 3. since the corrected O; variation index trend is incremented only for those sample collection periods that are determined to be noise free. Yes in Block 505, the present imention accounts for those periods when the O; variation indexes are likely associated with noise and adjusts the total Ch v ariation index trend accordingly by not incrementing

15 the corrected O; variation index trend when noise is likely i e , when the deviation is not less than the dev iation threshold. No in Block 505

!'^'Ki ^Q is a flow chart of a method of delivering a therapy in a medical dex ice according to an embodiment of the present invention FlG 10 is an exemplary graphical representation of an O; variation index trend utilized in a method of delivering a therapy in

20 a medical device according to an embodiment of the present invention Λs illustrated in

FIGS ^Q and I O, according to an embodiment of the present invention, once the first sample collection period, i e . 0-5 seconds, has expired, and therefore both the associated O; \ ariation index trend *?U0. Block 503. and the deviation of the associated samples from the trend. Block 504, o\ er that peπod have been determined and utilized to determine that

25 the sample is likely to be noise free. Yes in Block 505, the last trend value tθi, ι-» along the corresponding O> \ ariation index trend 000 is determined b> projecting the location of the last O2 variation index sample of the window Cb t-m onto the O; variation index trend QOO f^<or the initial sample collection period, the corrected O2 variation index trend cO;

30 is incremented in Block 5OS by being set equal to the last trended value tθi, ι_M. which is then utilized as the generated corrected O₂ variation index trend c(); foi the determination of Block 509 1 *1

The window of O^ sanation index samples shifts to include the next O? vacation index sample U2,jω and the previous a-\ O; variation index samples O₂^. ^₁ from the previous window υf samples The current O; variation index sampler O;,;.^ are then used to determine the next On variation index trend 902, with the value of the fast O; \ariation 5 index sample O^ u») being projected onto the current O2 variation index trend 902 to generate a corresponding last trended value iC>2_tu.v which is then utilized as the corrected O; variation index trend in the determination of Block 50*3, and so forth An O; v ariation index U end is then determined for the next window of O; variation index samples and the corresponding last trended v alue is determined by projecting the location of the last or

10 most recent O₂ variation index sample onto the O; variation index trend, and so forth

The process continues as described on a sample by sample basis until the effects of noise cause the O; \ ariation index samples to deviate from the O; \ ariation index trend so that the deviation becomes greater than the deviation threshold, No in Block 505 Once the deviation of the O; variation indexes is determined not to be less than the deviation

15 threshold, indicating a likelihood that the determined cardiac event is the iesuh of noise, a corrected O2 variation index offset, which operates to keep a amrang account of the non- noise ftee periods, is updated, Block 510 In particular, as the window of Oj variation index samples continues to be shifted to include a next O₂ variation index sample and the prc\ ious n-1 samples, and the corrected 0> variation index trend continues to be updated

20 accordingly, the leading edge of the window of O; variation index samples may begin to advance within a noise portion ^CX)6 Once the window of O; variation index samples adv ances far enough within the noise portion 906, the deviation of the samples in the current sample window will become greater than the deviation threshold. No in Block 505, and therefore the v alue of the corrected Oj \ ariation mde\ trend is held equal to the last

25 corrected Oi variation index trend that was not associated with noise

Assuming that the window of O> v ariation index samples that initially deviates from the corresponding O₂ variation index trend to be indicative of noise occurs at Oz v ariation index sample 0?o_>- which corresponds to trended O^ variation index sample tOi^v and the corrected Oi \ariaiion index is therefore no longer updated, the offset is

30 updated in Block 510 by being set equal to the difference between the cυπent trended 0> variation index ιθ?_(x, and the corrected O; variation index associated with the last window of samples determined to be noise free, illustrated in FIG I O by tθ_2co - cO_2cv.t , The updating of the offset continues, wtth the offset being updated m Block 510 for each 23 subsequent window of samples by continuing to determine the difference between the current trended O; \ariation index tθj^ _ml and the corrected Oi variation index associated with the last window of samples determined to be noise free, until another window of noise free O; \ ariation index samples is receh ed. Yes in Block 505 The updating of tine 5 offset in Block 510 as the w indow of samples advances during detection of the noise 906 can therefore be summarized general!) by the follow ing equation

offset tO;₍₁,_m> - cθ2_U_n Equation 4

10 where i is the first instance that noise is detected for a given noise period, m is the

Subsequent consecutive samples during this period of noise, and i-i corresponds to the last corrected Cb variation index trend \ aluo for which the associated window of samples was determined to be noise free immediately prior to the trend \alue for which the associated window of samples was determined to be likely corrupted by noise

15 Assuming the next noise free C); variation index sample is leceived for the O? variation index sample associated with O; variation index trend tO^o, for example, the incrementing of the corrected <>> \ ariation mde\ trend in Block S08 then continues and the updating of the offset in Block 510 is suspended As a result, the offset was last updated for the prc\ ious sample Q-M-nby being set equal to the difference between the value of the

20 On vari ation index trend tϋ;₀,,t₎ generated dun ng the v\ indow of samples occurring just prior to the initial noise free window of samples and the value of the corrected 0; variation index cOj^.n associated with the last window of samples determined to be noise free

The corrected O_? variation mde\ trend is then incremented in Block 508 by the

25 difference between the current noise free trend value tθ;,u and the sum of previously determined offsets, which in the example would be the offset associated with noise period 906. illustrated by tθ2κ-n - c€b _vi » it may also be noted that since noise period 906 is the first nois} period in the example, the offset equals the total change in the O^ variation index trend during the period of noise tUj^.}, - cθ;_cvι> The incrementing of the corrected

30 O; variation index trend continues with the window of O; variation index samples being shifted to include the next O; variation index sample O₂^ «■»_> and the previous 14 O; variation index samples starting from C> variation index sample O^ _Mιvu_>. so that the subsequent windows of samples are used to determine the next (h vacation index trends. 24 and the last C); variation index sample ϋ>₍ ,_m> ιs piqjected onto the current O^ sanation index trend to generate a corresponding trend \alue tOj,_Λ ,„, The corrected U₂ v ariation index trend is then incremented in Block 508 by subtracting the offset updated during the previous noise period 906 from the current trend \ aSue t():_u ,„-_> 5 Assuming, for example, that the current window Of O₂ variation index samples subsequently remains noise free for a noise fiee penod of time 908 and then deviates from the corresponding O; \ariation index trend, No in Block 505, to be indicative of a next noise portion 910 at O; variation πidex sample Ov>, which coπespoπds to tiended O₂ variation index sample tt^,. the corrected On variation index is no longer incremented for

10 trended O2 variation index sample tθ?_{(/ >}, while the offset is updated in Block 5 I 0 by being set equal to the difference between the current value of the non-noise free trended O2 variation index sample tθ:₍._> and the value of the corrected O; variation index associated with the last window of samples determined to be noise free, i e cO;_(/._{} !} The updating of the offset continues for each subsequent window of samples b\ taking the difference

15 between the value of the trended O; variation index generated for the current non-noise free window of samples and the value of the corrected O^ variation index associated with the last window of samples determined to be noise free, until the next noise fiee (h variation index sample is received

For example, assuming the ne\t noise free O; variation index sample is received

20 for the On variation index sample associated with trended Q2 variation index sample tθ:_uu- the incrementing of the corrected O; variation index trend in Block 508 then continues and updating of the offset in Block 510 is suspended As a result, the offset was last updated for the previous sample tO^-u b> being set equal to the difference between the "value of the O₂ variation index trend tθ^-n generated during the window of samples occurring

25 just prior to the initial noise free w irtdoxs of samples and the value of the corrected O. variation index associated with the last window of samples deteimined to be noise fiee, CO_XM , it should be noted that the offset at this point equals the sum of the changes observed during the two preceding mnoisy periods 906 and 910. [tO^-i r tθ;,, _\,] > [ tO;_(V

30 The corrected O₂ variation index trend is then computed as the difference between the cuiient noise free O; variation index U end \ alue tO^ and the cuirem offset in this wa\, the corrected On variation index trend is updated for subsequent windows of O; \ ariation index samples b) subtracting the sum of the pϊ evious changes in the (h vanation 25 index trend values associated with noise from the current O; variation index trend value, described generally by the following equation

cihn-n- i) ~ t();₀.n i) - 2_offset Equation 5

5 where i is the current O; variation index trend value and n is the number of samples in a sampling window Toffset represents the current offset at any point in time, where the summation, ^- '^s indicative of the fact that the current offset represents the sum of individual offsets accumulated during each individual period of noise since the start of the

10 flow chart As can be seen in Equation 5, the incrementing of the corrected O; variation index trend does not begin until the first a samples associated with the first window of samples 900 arc received

Λs illustrated in HG 10, each period of noise ^CXK>, 910 associated with, the O; variation index trend includes a start point 920 and an endpoint 922 Ideally, noise will be

15 detected, No in Block 505, when the endpoint of the associated Oj variation index tiend is located at or just beyond the start point 920 of the period of noise 906, ^c>10, and will subsequent!) no longer be detected, Yes m Block 505, w-hen the subsequent endpoint of the associated Cb \ariation index trend is located at or just beyond the endpoint 922 How ev er, as can be in the embodiment described above in reference to FlG I 0. when the

20 endpoints of the Cb variation index trends are used in the determination of both the incrementing of the corrected O; variation index trend. Block 508. and the updating of the offset, Block 510, the window of Oj variation index samples will be located bevond the start point 920 and within the noise portion when the noise is initially identified. No in Block 505. and beyond the endpoint 922 and outside the noise period when noise is

25 subsequently no longer detected. Yes in Block 505

According to an embodiment of the present invention, therefore, in order to increase the likelihood that the offset will correspond to the actual period of noise, the present invention utilizes a predetermined trend value other than the leading endpoint of the O; \ ariation index trend Rather than projecting on the first and the last sample of

30 each of the n samples in the samples of windows to obtain the first O; variation index trend value and the last O; variation index trend value for each geneiated O> valuation index trend as described abo\e. each sample within the window of samples is projected onto the ϋ; \ ariation uidex trend to αeneraie n Oj sanation index tiend values so that anv 26 one of the » end valuer can then be utilized during the incrementing and updating of the corrected Cb v ariation index trend and the offset, respective!}

For example, once the initial O; variation index trend ⁰OO has been computed. Block 503 and the delation of the samples from the On variation index trend 900 has been 5 determined. Block 504, resulting in a determination that the sample is not likely associated with noise. Yes in Block 505, the corrected (h variation index trend cO; is incremented in Block 508 b\ being set equal to a predetermined one of the 1 5 trend values located between the first trend value tθ;_{( t}) associated with the first Oj v ariation index sample of the window and the last trend value tt^i^ associated with the last O2 variation index

10 sample of the window O;_<j ^ According to an embodiment of the present invention, the corrected Cb variation index trend cO^ is incremented by being set equal to the central trend value, i e , Oz variation index trend value tθ;^> lhe process continues with the v\ indow of Oa variation index samples being shifted to include the next Oi variation index sample O;₍if>) and the previous 14 O^

15 variation index samples from the previous window of samples O;(2) through C);, 3 ^ The current window of samples O;^-IM is then used to determine the next O; variation index trend 902, and the cenua! (h variation index sample O^.is projected onto the current Oj variation index trend 902 to generate a corresponding central trend value tOj^ The corrected O; variation index trend is then incremented in Block 508 bv being set equal to

20 the central trend va!ue tOix, The process continues as described abo\ e using the predetermined trend value in place of the last O; variation index trend value to perform the incrementing and updating of the corrected O^ variation index' trend and the offset respectively

MG 1 1 is an exemplary graphical iepi escalation of generation of a corrected O;

25 variation index trend offset utilized in a method of delivering a therapy in a medical device according to an embodiment of the present in\ ention According to another embodiment of the present invention, rather than using a single trended value during the incrementing and updating of the corrected O> \ ariation index trend and the offset, respectively, multiple \ alues may be utilized l-^'or example, as illustrated in FIG i 1. the last or leading

30 end O; variation index trend is utilised during the incrementing of the O; variation index trend and a predetermined O? variation index U end \ alue is utilized during the updating of the offset The predetermined O; \ ariation index trend \ alue is chosen to increase the likelihood that the offset will correspond to the actual period of noise, such as the midpoint of the O; variation index trend, few example Jn this v\as , the incrementing of the corrected Oi v ariation index trend is performed using the leading endpoint of the determined O; variation index trend during the initial noise free period associated with O? variation index trends 90<λ 902, and so forth, and the process continues as described on a 5 sample by sample basis until the effects of noise cause the O; variation index samples to deviate ftom the Oj sanation index tiend, No in Block 505

Assuming again that the window of O; variation index samples that initial S\ deviates from the corresponding Oj \ ariatiυn index trend to be indicative of noise occurs at O; variation index sample O^ _o. which corresponds to O? variation index trend tO^j,

10 incrementing of the corrected O^ variation index is therefore suspended It should be noted that while computing an offset in Block 510. the corrected Cb variation index trend values belong to the immediately prior sampling window During the updating of the offset in Block 5 !0. the O; variation index trend offset is updated by determining the difference between the leading O. variation index trend value tO^v^ of the current noise

15 corrupted Oj variation index tiend tO^and the corrected C); \ ariauon index trend cθ>_tv<>₎ determined immediately prior to the central O; variation index trend tO^.s)

The piocess continues for subsequent w indows of samples, with the determination of whether the current window of samples are corrupted by noise being made in Block 505 and the updating of the offset in Block 510 being made based on the predetermined trend

20 value, until the next noise free O? \ ariation index sample is identified. Yes in Block 505, so that the offset is updated for each window of samples as set forth generally by the following equation

offset tθ;_t, _jn-d> - cθ2,i-n Equation ό

25 where i is the first instance that noise is detected for a given noise period, m is the next sample, d corresponds to the predetermined trend \ aiue associated with the current noise corrupted window of O; v ariation index samples, and i~l corresponds to the immediate last trend value associated with the window of samples determined to be noise free prior to

30 the predetermined trend value associated with the current noise corrupted window of samples

For example, if the next noise free O₂ variation index sample is received at O2 v ariation index trend tO^,, the offset was therefore last updated during the prev ious 28 trended O^ variation index sample !()>, _n by being set equal to the difference between the centra! O₂ variation index trend \ aiue tih^.ψ, of the current noise corrupted O; variation index trend and the corrected O; variation index trend eO;_o-9) determined prior to the centra! O? variation index trend t();^,_N, Incrementing of the corrected Cb variation index 5 trend cG^is resumed in Block 508 bv subtracting the offset 907 from the current O^

\ ariation index trend tO^,

The process continues during over m windows of Q; variation index samples occurring over the subsequent noise free portion 908, uith the corrected O; variation index being incremented, Block 508, by subtracting the offset 907 from the current trend value

10 tθ2_t\,_m) Assuming, for example, that the current window of Cb variation index samples subsequent^ remains noise free for a period of time associated with the noise free period 908 and then deviates from the corresponding O; variation index trend, No in Block 505, to be indicative of a next noise portion ^c>l 0 at O; variation index trend tO;,,^ incrementing of the corrected Ch v ariation index is therefore suspended and updating of the offset in

15 Block 5 S 0 resumes

During the updating of the offset, the O^ variation index trend offset is updated by determining the difference between the centra! Oj sanation index U end value tO^-x, of the current noise corrupted O; variation index trend tθ;_(/,and the corrected (> variation index trend cθ^^«>. determined immediately prior to the central O; \ ariation index trend tCht/.jo

20 The process continues for subsequent windows of samples, with the determination of whether the current window of samples are corrupted by noise being made in Block 505 and the updating of the offset in Block 510 being made based on the predetermined trend value using Equation 6, as in the previous noise period 906 until the next noise free O_^ variation index sample is identified. Yes in Block 505

25 For example, if the next noise free O; \ ariation index sample is received at O^ variation index trend tO;_{CΛ )}, the offset was therefore last updated for the previous O₂ variation index trend tO^-n by being set equal to the difference between the central O2 \ ariation index trend tθ₂,« y, associated with the pre\ ious O; variation index trend and the iast incremented corrected O; variation index trend cθ;_f,-_'j> prior to the central O; \ ariation

30 index trend cθ^_ft._in Incrementing of the corrected O₂ variation index trend cθ^_u,is resumed in Block 508 by subtracting the sum of the offsets 907 and ⁰OO from the current O; variation index trend tO^w, 29

The process continues during over ra window s of C); variation index samples occurring o\er the subsequent noise corrupted and noise free portions, with the offset being updated by determining the current offset using Equation o and the inciementing of the coπected O; \ariation index in Block 508 being performed by subtracting the sum of 5 the prior determined offsets from the current O; variation index trend, set forth generally in Equation S above

MG 12 is an exemplar} graphical representation of an O2 variation index trend utilised HI a method of delivering a therapy in a medical device according to an embodiment of the present invention FlG i 3 is a flow chart of a method of delivering a

10 therapy in a medical device according to an embodiment of the present invention As illustrated in HGS 12 and i3. according to an embodiment of the present indention, a /one 920 for identify ing when the event is most likely associated with an unstable hemodynamic event, such as v entricular taeh>eardia or ventricular fibrillation, is defined based on a non -physiologic event threshold 921 and a normal sinus rhythm threshold limit

15 922, so that once the C); \ ariaiion index trend is determined to be less than the baseline

\alue 210. a determination is made as to whether the O^ variation index trend is within the VT Vh /one 920, Block 513

In particular, for example, a determination is made as to whether the slope of the O; variation index trend, determined based on two of the known trended values of the O;

20 variation index trend, such as the first and the last trended value, for example, is either greater than the slope of threshold 921 or less than the slope of threshold ^c>22, and therefore outside the VTAT zone 920, Block 513 If the O? variation index trend is determined to be outside the VPVF zone, No in Block 513, the current stored slope "values are cleared and the determination as to whether the episode verification time period has

25 expired is made. Block 507. described above If the Ch \ ariation index trend is determined to be within the VT VF /one. Yes in Block 513. a determination is made as to whether the O; variation index trend is sustained, i e , remains within the 7 one 920 for a predetermined time period, such as over six samples 01 two seconds, for example. Block 515

30 If the O; variation index trend is not sustained, the process returns to Block 502 so that the deviation of the O? variation index samples from the O? variation index trend , Blocks 503 and 504 is determined for the next window of O: variation index samples, described above If the <)-> variation index trend is sustained within the VIVVF zone °20 30

for the predetermined period of time, the secondary confirmation process confirms the identification of the malignant cardiac event, and therapy is delivered. Block 512. and the cυπent stored slope values aie cleared

Both threshold 92 U which corresponds to abrupt changes in the slope of the ih 5 variation index trend indicative of non-physiological events, such as a change in posture for example, and threshold 922, which corresponds to normal sinus rhythm, are programmable According to an exemplar} embodiment of the present invention, threshold 021 corresponds to the Oa variation index trend crossing the baseline value 210 in five seconds or less, so that threshold 921 corresponds to a slope of 0 004 (i c , 0 02

10 div ided by 5 seconds), and threshold 922 corresponds to the Ch variation index trend crossing the baseline \alue 210 in 20 or more seconds, so that threshold 922 corresponds to a slope of 0 001 (i o . 0 02 divided by 20 seconds)

FlG !4 is an exemplary graphical representation of an O; variation index, trend utilized in a method of delivering a therapv in a medical device according to an

15 embodiment of the present im eπtiors FlG 15 is a flow chart of a method of delivering a therapy tn a medical device according to an embodiment of the present in\ ention As illustrated in FIGS 14 and ] 5, a fast YT threshold 923 is included vathin the VT/YF /one 920 in order to discriminate YT from VF events, with the event being identified as a fast YT event when the On variation index trend is located between threshold 922 and

20 threshold 923, and as a VF e\ent when the Oi variation index trend is located between threshold 921 and threshold 923

According to another embodiment of the present invention, a slow Y 1 threshold 925 may also be included in order to discriminate between normal sinus rhythm and slow VT events, with Oa variation index trends that aie sustained between threshold 925 and

25 threshold 922 being identified as associated with a slow- VT event

In particular, once the O? variation index trend is determined to be less than the baseline value 210, Yes in Block 509, and the O₂ variation index trend is determined to be within the YT/YF /one 920. Yes in Block 513,, as described above, a determination is made as to whether the O; variation index trend is determined to be sustained within the

30 VT/VF /one 920 for a predetermined time period. Block 5 S ^. such as 3-5 samples, for example If the O; variation index trend is not sustained VT VF, No in Block S S 5. the process returns to Block 502 so that the deviation of the O₂ variation index samples from 3 1 the <>₂ variation index trend , Blocks 503 and 504 is detei mined for the nexi window of O> variation index samples, described abo\e

If the O; variation index trend is sustained VT/VF, Yes in Block 515, the event is identified as a VF event if the O; variation index trend is located between threshold ^c>21 5 and threshold 923, and as a VT event if the O₂ variation index trend is located between threshold 923 and threshold °22, Block 517 Once the classification of the event ts determined in Block 517, delivery of the therap\ is adjusted accordingly. Block 519, and the cuπent stored slope values are cleared

According to an embodiment of the present invention, slow VT threshold 925 may

10 also be included in order to discriminate between normal sinus rhythm and slow VT events in particular, if the O: variation index trend is not determined to be within the VT/VF /one. No in Block 513, a determination is made as to whether the O; variation index trend is located between threshold 925 and threshold °22 If the O; \ ariation index trend is located between threshold ^c>25 and threshold 922, the event is classified as a slow

15 VT event and the classification may be stored for future reference

Similar to threshold 92 i and threshold 922. both threshold 923 and threshold 925 are programmable According to an exemplary embodiment of the present invention threshold 923 corresponds to the O₂ variation index trend crossing the baseline \alue 210 in twelve seconds or less, so that threshold °23 corresponds to a slope of 0 0017 (i e , 0 02

20 divided by 12 seconds), and threshold °25 corresponds to the O2 variation index trend crossing the baseline value 210 in 28 or more seconds, so that threshold 925 corresponds to a slope of 0 0007 (i e , 0 02 divided bv 28 seconds)

As described above, when noise is detected, a correction can be made by referring to the coπected O; variation index trend at a pπoi instant when the O; variation indc\

25 trend was unaffected by noise Since the determination of noise by measuring deviation of the O? v ariation index samples from the O? variation index trend is done over a window of multiple samples it may be necessary to look back substantially more than a single, immediately prior sample, which may lead to a delay in the determination of the presence of noise from its actual moment of onset According to an embodiment of the present

30 invention, in order to mitigate any error due to such a delay, the corrected O> variation index U end consists of multiple values each corresponding to one sample within the sampling window for the determination of the O2 variation index trend, such that when noise is detected in a sampling window, an early value from a prev iously determined 32 group of values of the corrected O; variation index trend can be ief erred to as one belonging to a noise-free period This particular value of the corrected O; variation index trend can then be used for computing an offset to correct further On variation index trend In this embodiment for each sampling window, a group of corrected ih variation index 5 trend values are computed and stored in device 14, each value corresponding to one sample of the On variation index within the window The method of computation of the value of the corrected 0_ variation index trend depends on the computed deviation during the sampling window. Block 505 of FlG 9, and, in case of a Uuge deviation indicative of noise. No in Block 505 of FIG 9, all the corrected O^ variation index trend values for the

10 sampling window arc set to the value of the corrected Cb variation index trend corresponding to a predetermined sample of a predetermined prior sampling window such as the first sample, labeled /-1 , of the sampling window In case of a largo dc\ iation, No in Block 505 of !¹IG °, the offset is also updated in Block 510 by assigning to it the difference between the value of the O; variation index trend corresponding to a

15 predetermined sample, such as the sample labeled / n of the sampling window, and the value of the corrected O2 variation index trend corresponding to a predetermined sample of a predetermined prior sampling window, such as the fust sample, labeled /-1 of the sampling w indow Such computation may be continued for ever\ sample of the O₂ \ ariation indev trend where the deviation is laige, Ko in Block 505 of FlG ^c>. according to

20 the equation

offset ^~ K h,_{n %} - rfλ^\ Equation 7

where fO.-, , , is the O2 variation index trend value corresponding to the predetermined 25 sample m within the current sampling window and c();,S' is the value of the corrected O; variation index trend corresponding to the predetermined sample k within a prior sampling window that precedes the cuirem sampling window by a predetermined number of window s indicated by p The \ alue of the offset so computed is used during the next noise- free sampling window 30 In case of a small de\ iation. Yes in 505 of FiG 9. indicative of a noise-free sampling window, the corrected CK variation index trend is computed corresponding to each sample 33 within the interval by taking the difference between the C); variation index trend corresponding to each sample and the offset computed during the last period of noise Since this method depends on the corrected On variation index trend from a prior sampling window, it requires initialization of the corrected ih variation index trend for a 5 certain number of initial sampling v\ indow s, the number being same as the predetermined

\ aloe, p For each of these window s if the deviation of the O? variation index samples from its O; variation index trend is above its threshold for the determination of noise, each value of the corrected O; variation index trend in that sampling window is set to zero otherwise each v alue of the corrected CK v ariation index trend is set to its corresponding 10 value of the O; variation index trend

Determination of the hemodynamic status raav be improved bv further analyzing the slope of the Cb variation index trend The slope being referred to here is the slope of the U₂ \ aria ti on index trend line with the time axis w hich ma> also be referred to as the rate of change of the O^ variation index trend Such analysis of the slope will enable 15 determination of the degree of the perfusion loss which may van, between tolerated

\entricuJar tachyarrhythmia, non-tolerated \entricular tachyarrhythmia and \entncular fibrillation A slope is computed for each sampling window, of 5 second duration consisting of i 5 samples for example, and a certain number of the most recent values of it are stored in the device memory The slope mav be defined as the difference between the 20 first and the last O₂ variation index trend values, for example tih{\ <Λ - tO^i > for the sampling window consisting of the T¹ through the 15^lh sample It may also be defined as the ratio of the difference between the first and the last CK variation index trend values and the time interval between them For yet another definition of the slope, the difference and the ratio defined above may be computed over a plurality of subsections within a sampling 25 window and can be further combined to derive a composite slope parameter A plurality' of the slope values so computed are stored in the device memory According to an embodiment of the present invention, the range of v alues of the slope of the corrected O; variation index trend is divided into predetermined groups corresponding to \ arious cardiac rh\ thras and a predetermined group of v alues considered non- 30 physiologic For example a signed slope value smaller than -O 007 per second, called the physiologic limit, may be considered non-physiologic and any signed slope value larger than -0 00007 per second, called the sinus limit, mav be considered to correspond to a heniodynamically stable, benign cardiac T hythm Any intei mediate slope value betvv een 34 these two limits ina\ be considered to correspond to YT or VF Alternative!) , the range of those intermediate slope values may be further subdivided with a hemodynamic stability limit demarcating the boursdarv, separating a hemodynamically stable YT, also called the tolerated VT, from a hemoch namically unstable VI. also called the non-tolerated VI, and 5 YF The physiologic, sinus and the hemodynamic stability limits also mav be patient- specific and determined based on tests such as defibrillation threshold test and other electrophysiologic tests

The slope of the corrected O; variation index trend is compared against the piedeterrmned physiologic, sinus and the hemodynamic stability limits to determine the underlying

10 cardiac rhythm and the hcmod) narnic status of the patient, such as a sinus rhythm, a stable

V I , an unstable YT or YF or a non-physiologic signal If it is determined that the corrected Cb variation index trend does not correspond to VT or VF. No in Block 51 > of FICJ 13, an> \alue of the slope stored in the dev ice 14 is removed, and a determination is made as to whether a predetermined episode verification time period has expired, Block

15 507 if ihe episode verification time peiiod has not expired, the process returns to Block

502 so that the deviation of Cb variation indexes from the O; v ariation index trend. Blocks

503 and 504, is detei mined for ihe next window of <);> variation indexes (>00 If on the other hand it is determined that the corrected Cb variation index trend corresponds to VT or YF, Yes in Block 513 of FlG 13, a determination is also made as to

20 whether it is sustained by collecting additional samples of Cb \ ariation index, Block 502 in

FlG 13, and repeating the subsequent steps for a predetermined number of samples, for example 6, or a predetermined duration, for example 2 seconds, to arrive at the same conclusion. Yes in Block 513 Ihe slope value from each pass through the flow chart during such attainability determination, 515 in Hgure 13, is stored HI the device 14

25 The consistency of the corrected O; variation index trend slope is determined b\ computing the difference between the maximum and the minimum values of the slopes or the standard de\ iation or the variance of the values of the slopes If such values are smaller than a predetermined limit, the rhythm is classified in Block 517 based on the average or the minimum signed value of the slope as either VT and YK or further

30 classified as stable VT oi unstable VT and VF and a therapy is determined based on it,

Block 5 f ° The stored slope values ate removed from the device subsequently. Block 523

While a particular embodiment of the present invention has been shown and described, modifications may be made It is therefore intended in ihe appended claims to 35 cover a!! such changes and modifications, which fall within the true spirit and scope of the invention.

Claims

36

We Ci aim:

! A medical device, comprising a plurality of sensors sensing cardiac signals.

5 a physiologic sensor sensing physiologic signals to generate a plurality of variation index samples corresponding to the sensed signals, and a microprocessor detecting a cardiac event in response to the sensed cardiac signals, computing a variation index trend associated with a predeteimϊned number of variation index samples of the plurality of v ariation index samples, determining 10 a rate of change of the variation index trend, and confirm) ng the detenni ned cardiac event in response to the computed variation index trend and the determined rate of change

2 The dc\ ice of claim i, w herein the variation index trend corresponds to a measure 15 of change i n ti asue oxygen ati on

3 The device of claim I , wherein the microprocessor determines a dev iation of the predetermined number of variation index samples from the v ariation index trend

20 4 The device of claim 3, w herein the microprocessor determines whether the computed variation index trend is less than a predetermined index trend threshold

5 I he device of claim i, wherein the microprocessor determines whether the rate of change is one of greatei than an upper limit and less than a lowei limit 25 b The device of claim I, wherein the microprocessor whether the rate of change is within one of a first range corresponding to a v entricular tachycardia event and a second range corresponding to a ventricular fibrillation event

30 7 The device of claim S, wherein the variation index trend corresponds to relationships between proportions of red light intensities and infrared light intensities to corresponding baseline intensities

8 The device of claim 1 , wherein the variation index trend corresponds to

_^ wherein ? is an intensity of red light, /^" is an intensih of mirared light, /,. is an intensity of red light baseline, and /^* is an intensity of infrared light baseline

9 A method of determining a cardiac event in a medical device, comprising' sensing cardiac signals from a plurality of electrodes. determining the cardiac event in response to the sensed first cardiac signals, sensing physiologic signals and acquiring a plurality of variation index samples

10 corresponding to the sensed physiologic signals, computing a variation index trend associated with a predetermined number of variation index samples of the plurality of variation index samples, determining a rate of change of the computed v ariation index trend, and confirming the determined cardiac event in response to the computed variation

15 index trend and the determined rate of change

10 i^'he method of claim 9, wherein the variation index trend corresponds to a measure of change in tissue oxygenation

20 I I The method of claim 9. further comprising determining a deviation of the piedetermtned number of variation index samples from the variation index, trend

12 The method of claim I S, further comprising determining, in response to the determined deviation, whether the sensed cardiac signals are associated with an unstable

25 fhythm, v\ herein the determined cardiac event is confirmed in response to the determined deviation and the sensed cardiac signals being associated with an unstable rhythm

13 The method of claim 9 wherein the variation index U end coπesponds to relationships between proportions of red light intensities and infrared light intensities to

30 corresponding baseline intensities 3 S

14. The method of claim 9. wherein the variation index trend corresponds to — — ~ , wherein i is an intensity of red light, f is an intensity of infrared light, /,. is an intensity of red light baseline, and C is an intensity of infrared light baseline. 5

15. The method of claim 9, further comprising determining whether the rate of change is one of greater than an upper limit and less than a lower limit.

16. The method of claim 9, further comprising determining whether the rate of change 10 is within one of a first range corresponding to a ventricular tachycardia event and a second range corresponding to a ventricular fibrillation event.

17. A method of verifying a cardiac event in a medical device, comprising: sensing physiologic signals and acquiring a plurality of variation index samples

15 corresponding to the sensed physiological signals; computing a variation index trend associated with a predetermined number of variation index samples of the plurality of variation index samples; determining a rate of change of the variation index trend; and verifying the cardiac event in response to the determined variation index trend and

20 the determined rate of change.

18. The method of claim 17, wherein the variation index trend corresponds to a measure of change in tissue oxygenation.

25 19 The method of claim 17, further comprising determining whether the computed variation index trend is less than a predetermined index trend threshold.

20. The method of claim 19, further comprising determining whether a deviation of the predetermined number of variation index samples from the computed variation index trend 30 is greater than a deviation threshold, wherein the cardiac event is verified in response to both the computed variation index trend being less than the predetermined index trend threshold and the deviation not being greater than the deviation threshold. 3<⁾

21 The method of claim 20, wherein the variation index trend corresponds to relationships between proportions of red light intensities and infrared light intensities to corresponding baseline intensities.

5

22 The method of claim 17, wherein the variation index trend corresponds to J^{" ~ "}.*^' , wherein ; is an intensity of red light, f is an intensity of infrared light, /,, is art intensity of red light baseline, and /^* is an intensity of infrared light baseline. 10

23 The method of claim 17, further comprising determining whether the rate of change is one of greater than a.n upper limit and less tfmn a lower limit

24. The method of claim 17, further comprising determining whether the rate of 15 change is within one of a first range corresponding to a ventricular tachycardia event and a second range corresponding to a ventricular fibrillation event.