Rate responsive heart stimulation device using neural network and IEGM classifier

• US 5,782,885 A
• Filed: 02/10/1997
• Issued: 07/21/1998
• Est. Priority Date: 02/12/1996
• Status: Expired due to Term

• Current Assignee: Pacesetter AB
• Original Assignee: Pacesetter AB
• Inventors: Andersson, Jonas
• Primary Examiner(s): Kamm, William E.
• Assistant Examiner(s): LAYNO, CARL HERNANDZ
• Application Number:
  US08/797,422
• Time in Patent Office: 526 Days
• Field of Search: 607/9, 607/11, 607/17, 607/25, 600/523, 600/525
• US Class Current: 607/17
• CPC Class Codes: A61N 1/365 controlled by a physiological parameter, e.g. heart potential evoked response A61N1/371

Abstract

In a method for classifying IEGM waveforms in dependence of the workload of a patient, a predetermined number of IEGM-signals, each signal extending over at least one segment of one heart beat cycle, are registered, whereafter the IEGM-signals are fed to a neural network and an encoded form of the signals is formed using said neural network. This encoded form is stored in a memory, for use in classifying further registered IEGM-signals. In an active cardiac implant connectable to an implantable electrode arrangement adapted for in vivo delivery of stimulation pulses to a heart, IEGM signals present are obtained from one or more of the electrodes. A pulse generator connected to the electrode arrangement, generates and emits stimulation pulses with a variable stimulation interval between successive stimulation pulses. The implant also contains a classifying device for classification of a predetermined number of IEGM-signals registered during predetermined time intervals at predetermined points of time according to a predetermined classification stored in the classification device, this classification being related to preregistered waveforms of measured IEGM-signals. A control unit in the implant supplies a control signal to a control input of the pulse generator dependent on the classification of each of the registered IEGM signals. The control signal causes the pulse generator to adjust the stimulation rate dependent on each of the registered IEGM-signals.

Claims

1. A method for setting a cardiac stimulation rate dependent on a workload experienced by a patient, comprising the steps of:

obtaining a predetermined number of IEGM signals from a patient respectively while said patient is subjected to different workloads, each IEGM signal having an IEGM waveform associated therewith extending over at least one segment of one heartbeat cycle of said patient;

supplying said IEGM signals to a neural network and forming an encoded form of said IEGM waveforms using said neural network;

storing said encoded form of said IEGM waveforms in a memory;

obtaining subsequent IEGM signals, each subsequent signal having an IEGM waveform associated therewith, and using said IEGM waveforms in encoded form stored in said memory for classifying the respective waveforms of said subsequent IEGM signals to identify a current workload of said patient; and

administering pacing stimulation pulses to said patient at a stimulation rate matched to said current workload of said patient.

2. A method as claimed in claim 1 wherein the step of obtaining said IEGM signals comprises, for each workload, pacing said patient at a plurality of different stimulation rates and obtaining an IEGM signal with an IEGM waveform for each stimulation rate for each workload so that each stimulation rate for each workload has an IEGM waveform associated therewith, which is subsequently formed in encoded from by said neural network, and storing said stimulation rates respectively correlated with their associated IEGM waveforms in encoded form.

3. A method as claimed in claim 2 comprising the additional step of supplying stimulation pulses at a current stimulation rate to said patient while obtaining said subsequent IEGM signals, and wherein the step of using IEGM waveforms for classifying comprises, for each subsequent IEGM signal, classifying said subsequent IEGM signal by matching its IEGM waveform to one of the stored IEGM waveforms and identifying the stimulation rate associated with said one of said stored waveforms, and said method comprising the additional step of changing said current stimulation rate to the stimulation rate associated with said one of said stored IEGM waveforms if said current stimulation rate is different from said stimulation rate associated with said one of said stored IEGM waveforms.

4. A method as claimed in claim 1 wherein the step of forming an encoded form of said IEGM waveforms using said neural network comprises selecting at least one segment of said heartbeat cycle represented by an IEGM signal and using said at least one segment for forming said encoded form of the IEGM waveform associated with that IEGM signal.

5. A method as claimed in claim 1 wherein the step of forming an encoded form of said IEGM waveforms using said neural network comprises obtaining an average of a plurality of said at least one segment of each heartbeat cycle over a plurality of heartbeat cycles respectively represented by IEGM signals and using said average for forming said encoded form of the IEGM waveforms respectively associated with those IEGM signals.

6. A method as claimed in claim 1 wherein the step of forming an encoded form of said IEGM waveforms using said neural network comprises obtaining an average of a plurality of heartbeat cycles respectively represented by IEGM signals and using said average for forming said encoded form of the IEGM waveforms respectively associated with those IEGM signals.

7. A method as claimed in claim 1 wherein the step of obtaining a plurality of IEGM signals comprises obtaining a plurality of IEGM signals respectively from a plurality of consecutive heartbeat cycles.

8. A cardiac assist system for pacing a heart of a patient dependent on a current workload experienced by said patient, said system comprising:

a housing having a size and shape adapted for implantation in a patient;

pulse generator means, disposed in said housing, for generating and emitting stimulation pulses with a variable stimulation interval between successive stimulation pulses, said pulse generator means having an output at which said stimulation pulses are present, and a control input;

electrode means, connectable to said output of said pulse generator means, for delivering said stimulation pulses in vivo to the heart of said patient;

monitoring means, contained in said housing and connected to said electrode means, for registering IEGM signals associated with said heart at different workloads and different stimulation intervals, each IEGM signal having an IEGM waveform associated therewith;

memory means in said housing for storing a plurality of IEGM waveforms respectively associated with said IEGM signals at said different workloads and different stimulation intervals;

classifier means, having access to said memory means and supplied with subsequent IEGM signals from said monitoring means, respectively obtained at predetermined points in time during a predetermined time interval, for classifying a predetermined number of said subsequent IEGM signals by classifying the IEGM waveforms respectively associated with said subsequent IEGM signals as matching an IEGM waveform in said memory means; and

control means, contained in said housing and connected to said classifier means, for generating a control signal dependent on the classification of said subsequent IEGM signals and for supplying said control signal to said control input of said pulse generator means for varying the stimulation interval dependent on the workload associated with the subsequent IEGM signals.

9. A cardiac assist system as claimed in claim 8 wherein said classifier means comprises a neural network, contained in said housing, and supplied with at least one segment of one heartbeat cycle for each IEGM signal, said neural network forming an encoded form of the IEGM waveform associated with said at least one segment of each IEGM signal, and wherein said memory means comprises means for storing said encoded form of each IEGM waveform.

10. A cardiac assist system as claimed in claim 8 wherein said classifier means comprises a neural network disposed externally of said housing, said neural network being supplied with a predetermined number of said IEGM signals, each IEGM signal extending over at least one segment of one heartbeat cycle, said neural network forming encoded forms of the respective IEGM waveforms associated with each IEGM signal, and said system further comprising telemetry means for telemetrically transmitting said predetermined number of IEGM signals from said housing to said neural network and for telemetrically receiving said IEGM signals in encoded form and for supplying said IEGM signals in encoded form, upon receipt thereof, to said memory means for storage therein.

11. A cardiac assist system as claimed in claim 8 wherein said classifier means comprises a fuzzy logic processor disposed externally of said housing, said fuzzy logic processor being supplied with a predetermined number of said IEGM signals, each IEGM signal extending over at least one segment of one heartbeat cycle, said fuzzy logic processor forming encoded forms of the respective IEGM waveforms associated with each IEGM signal, and said system further comprising telemetry means for telemetrically transmitting said predetermined number of IEGM signals from said housing to said fuzzy logic processor and for telemetrically receiving said IEGM signals in encoded form and for supplying said IEGM signals in encoded form, upon receipt thereof, to said memory means for storage therein.

12. A cardiac assist system as claimed in claim 8 wherein said memory means comprises means for storing a plurality of IEGM waveforms respectively obtained while pacing said patient at a plurality of different stimulation rates so that each stimulation rate for each workload has an IEGM waveform associated therewith with said stimulation rates respectively correlated in said memory means with their associated IEGM waveforms, and wherein said classifier means comprises means for classifying said current IEGM signal by matching its IEGM waveform to one of the stored IEGM waveforms and identifying the stimulation rate associated with said one of said stored waveforms, and wherein said control means comprises means for changing said stimulation interval to a stimulation interval associated with the stimulation rate of said one of said stored IEGM waveforms if said current stimulation rate is different from the stimulation rate associated with said one of said stored IEGM waveforms.

13. A cardiac assist system as claimed in claim 8 wherein said classifier means comprises means for classifying respective segments of the IEGM signal for each heartbeat cycle occurring between 175 ms and 425 ms after generation of a stimulation pulse by said pulse generator means.