Asystole detection for cardiopulmonary resuscitation
A patient monitoring device ( 20 ) employing an ECG monitor ( 24 ) and a controller ( 26 ). In operation, the ECG monitor ( 24 ) monitors a corrupted ECG waveform ( 30 ), and the controller ( 26 ) classifies the corrupted ECG waveform ( 30 ) as one of a non-shockable rhythm or a potentially shockable rhythm. The corrupted ECG waveform ( 30 ) is classified by the controller ( 26 ) as the non-shockable rhythm responsive to a detection by the controller ( 26 ) of a presence of an asystole rhythm within the corrupted ECG waveform ( 30 ). Conversely, the corrupted ECG waveform ( 30 ) is classified by the controller ( 26 ) as the potentially shockable rhythm responsive to a detection by the controller ( 26 ) of an absence of the asystole rhythm within the corrupted ECG waveform ( 30 ) or an indetermination by the controller ( 26 ) as to the presence of the asystole rhythm within the corrupted ECG waveform ( 30 ).
1. A patient monitoring device, comprising:
an ECG monitor operable to monitor a corrupted ECG waveform; and
a controller operably connected to the ECG monitor to classify the corrupted ECG waveform as one of a non-shockable asystole rhythm, or a potentially shockable non-asystole rhythm,
wherein the controller is configured to detect a presence or an absence of an asystole rhythm within a segment of the corrupted ECG waveform as indicated by at least one time domain feature and at least one frequency domain feature extracted by the controller from the segment of the corrupted ECG waveform,
wherein the controller is further configured to classify the corrupted ECG waveform as the non-shockable asystole rhythm responsive to a detection by the controller of the presence of the asystole rhythm within the corrupted ECG waveform, and
wherein the controller is further configured to classify the corrupted ECG waveform as the potentially shockable non-asystole rhythm responsive to a detection by the controller of the absence of the asystole rhythm within the corrupted ECG waveform.
2. The patient monitoring device of claim 1 , wherein the controller is further configured to classify the corrupted ECG waveform as the potentially shockable non-asystole rhythm responsive to an indetermination by the controller as to the presence or the absence of the asystole rhythm within the corrupted ECG waveform.
3. The patient monitoring device of claim 1 , wherein the patient monitoring device is at least one of an automatic external defibrillator or an advanced life support defibrillator.
4. The patient monitoring device of claim 1 , wherein the patient monitoring device comprises at least one of an automatic external defibrillator, an advanced life support defibrillator or a patient monitor.
5. The patient monitoring device of claim 1 , wherein the controller includes:
a signal conditioner operably connected to the ECG monitor and configured to condition the segment of the corrupted ECG waveform; and
a feature extractor operably connected to the signal conditioner and configured to extract the at least one time domain feature and the at least one frequency domain feature from the conditioned segment of the corrupted ECG waveform.
6. The patient monitoring device of claim 5 , wherein the signal conditioner removes any artifact from the segment of the ECG waveform.
7. The patient monitoring device of claim 5 , wherein the signal conditioner removes a chest compression artifact from segment of the ECG waveform.
8. The patient monitoring device of claim 5 , wherein the controller further includes a rhythm classifier operably connected to the feature extractor and configured to detect any indication by the at least one time domain feature and the at least one frequency domain feature of the presence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform.
9. The patient monitoring device of claim 5 , wherein the controller further includes a rhythm classifier operably connected to the feature extractor and configured to detect any indication by the at least one time domain feature and the at least one frequency domain feature of the absence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform.
10. The patient monitoring device of claim 5 , wherein the controller further includes a rhythm classifier operably connected to the feature extractor and configured to detect any failure by the at least one time domain feature and the at least one frequency domain feature to indicate one of the presence or the absence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform.
11. The patient monitoring device of claim 1 , wherein the at least frequency domain feature includes an amplitude spectrum area and the at least one time domain feature includes a first difference power of the corrupted ECG waveform.
12. The patient monitoring device of claim 1 , wherein the controller being configured to detect the presence or the absence of the asystole rhythm within the segment of the corrupted ECG waveform as indicated by the at least one frequency domain feature includes the controller configured to:
extract an amplitude spectrum area based frequency variable from the segment of the corrupted ECG waveform; and
compare the amplitude spectrum area based frequency variable to at least one of an asystole detection threshold and a not-asystole detection threshold.
13. The patient monitoring device of claim 1 , wherein the controller being configured to detect the presence or the absence of the asystole rhythm within the segment of the corrupted ECG waveform as indicated by the at least one time domain feature includes the controller configured to:
extract a first difference power based time variable from the segment of the corrupted ECG waveform; and
compare the first difference power based time variable to at least one of an asystole detection threshold and a not-asystole detection threshold.
14. A patient monitoring device controller, comprising:
a feature extractor configured to extract at least one time domain feature and at least one frequency domain feature from a segment of a corrupted ECG waveform,
an asystole advisor operably connected to the feature extractor to classify the corrupted ECG waveform as one of a non-shockable asystole rhythm or a potentially shockable non-asystole rhythm based on an extraction by the feature extractor of the at least one time domain feature and at least one frequency domain feature from the segment of the corrupted ECG waveform,
wherein the asystole advisor is configured to detect a presence or an absence of an asystole rhythm within the segment of the corrupted ECG waveform as indicated by the least one time domain feature and at least one frequency domain feature,
wherein the asystole advisor is further configured to classify the corrupted ECG waveform as the non-shockable asystole rhythm responsive to a detection by the asystole advisor of the presence of the asystole rhythm within the corrupted ECG waveform, and
wherein the asystole advisor is further classified to classify the corrupted ECG waveform as the potentially shockable non-asystole rhythm responsive to a detection by the asystole advisor of the absence of the asystole rhythm within the corrupted ECG waveform; and
a shock advisor operably connected to the asystole advisor to classify the corrupted ECG waveform as either a shockable non-asystole rhythm or a non-shockable non-asystole rhythm responsive to the asystole advisor classifying the corrupted ECG waveform as the potentially shockable non-asystole rhythm.
15. The patient monitoring device controller of claim 14 , wherein the asystole advisor classifies the corrupted ECG waveform as the potentially shockable non-asystole rhythm responsive to an indetermination by the asystole advisor as to the presence or the absence of the asystole rhythm within the corrupted ECG waveform.
16. The patient monitoring device controller of claim 14 , further comprising:
a signal conditioner operable to condition the segment of the corrupted ECG waveform; and
wherein the feature extractor is operably connected to the signal conditioner and configured to extract the at least one time domain feature and at the at least one frequency domain feature from a conditioned segment of the corrupted ECG waveform.
17. The patient monitoring device controller of claim 16 , wherein the asystole advisor includes:
a rhythm classifier operably connected to the feature extractor and configured to:
(i) detect any indication by at least one of the at least one time domain feature and the at least one frequency domain feature of the presence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform,
(ii) detect any indication by the at least one of the at least one time domain feature and the at least one frequency domain feature of the absence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform, and
(iii) detect any failure by the at least one of the at least one time domain feature and the at least one frequency domain feature to indicate one of the presence and the absence of the asystole rhythm within the conditioned segment of the corrupted ECG waveform.
18. The patient monitoring device controller of claim 14 , wherein the at least frequency domain feature includes an amplitude spectrum area and the at least one time domain feature includes a first difference power of the corrupted ECG waveform.
19. The patient monitoring device controller of claim 14 ,
wherein the feature extractor being configured to extract at least one frequency domain feature the segment of the corrupted ECG waveform includes the feature extractor configured to extract an amplitude spectrum area based frequency variable from the segment of the corrupted ECG waveform; and
wherein the asystole advisor being configured to detect the presence or the absence of the asystole rhythm within the segment of the corrupted ECG waveform as indicated by the least one frequency domain feature includes the asystole advisor configured to compare the amplitude spectrum area based frequency variable to at least one of an asystole detection threshold and a not-asystole detection threshold.
20. The patient monitoring device controller of claim 14 ,
wherein the feature extractor being configured to extract at least one time domain feature the segment of the corrupted ECG waveform includes the feature extractor configured to extract a first difference power based time variable from the segment of the corrupted ECG waveform; and
wherein the asystole advisor being configured to detect the presence or the absence of the asystole rhythm within the segment of the corrupted ECG waveform as indicated by the least one time domain feature includes the asystole advisor configured to compare the first difference power based time variable to at least one of an asystole detection threshold and a not-asystole detection threshold.