# US 12,734,367 B2 · Techniques for predicting and treating refractory ventricular fibrillation

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,734,367 B2
- Document ID: US12734367B2
- Status: Granted
- Kind: B2
- Application number: 19/126,999
- Filed: 2023-10-30
- Published: 2026-01-01
- Granted: 2026-09-15
- Earliest claimed: 2022-11-02
- Assignees: University of Washington
- Applicants: University of Washington
- Inventors: Jason Coult; Thomas Rea; Peter J. Kudenchuk; Heemun Kwok; Jose Nathan Kutz
- Web page: https://patentgrep.com/patent/US12734367B2

## CPC

- A61N 1/3925: {Monitoring; Protecting}
- A61B 5/361: Detecting fibrillation
- A61B 5/4836: {Diagnosis combined with treatment in closed-loop systems or methods (A61B5/0036 takes precedence)}
- A61B 5/4848: {Monitoring or testing the effects of treatment, e.g. of medication}
- A61B 5/726: {using Wavelet transforms}
- A61B 5/7264: {Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems}
- A61N 1/39044: {in combination with cardiopulmonary resuscitation [CPR] therapy}
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- A61H 31/005: {with feedback for the user}
- G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients
- A61N: ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY (measurement of bioelectric currents A61B; surgical instruments, devices or methods for transferring non-mechanical forms of energy to or from the body A61B18/00; anaesthetic apparatus in general A61M; incandescent lamps H01K; infrared radiators for heating H05B)
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- A61H: PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY (electrotherapy, magnetotherapy, radiation therapy or ultrasound therapy A61N)

## Abstract

An example method includes identifying a first segment of an electrocardiogram (ECG) that ends prior to administration of an electrical shock and identifying at least one second segment of the ECG that begins after the administration of the electrical shock. The example method further includes predicting, based on the first segment and the at least one second segment, that the ECG is indicative of a defibrillation-resistant heart rhythm. The first segment is converted into a first scalogram, and the at least one segment is converted into at least one second scalogram. The defibrillation-resistant heart rhythm is predicted based on comparisons between the scalograms and the eigenscalograms. An indication of the predicted defibrillation-resistant heart rhythm is output.

## Claims

### Claim 1

A medical device, comprising:
at least one sensor configured to detect an electrocardiogram (ECG) of a patient who is in a state of cardiac arrest;
an output device configured to output an instruction to administer a treatment; and
a processor configured to:
identify a pre-shock segment of the ECG, the pre-shock segment being detected by the sensor when a heart of the patient is in a state of ventricular fibrillation (VF) during a time interval preceding administration of an initial electrical shock to the heart of the patient;
identify at least one post-shock segment of the ECG, the at least one post-shock segment being detected by the sensor during a time interval following administration of the initial electrical shock to the heart of the patient;
predict, based on the pre-shock segment and the at least one post-shock segment, that the patient is exhibiting shock-refractory VF by:
converting the pre-shock segment into a first scalogram;
converting the at least one post-shock segment into at least one second scalogram;
generating comparisons between the first scalogram and the at least one second scalogram to eigenscalograms; and
determining whether the comparisons are indicative of shock-refractory VF; and
in response to predicting that the patient is exhibiting shock-refractory VF, cause the output device to output the instruction to administer a treatment.

### Claim 2 (depends on 1)

The medical device of claim 1, wherein the processor is configured to convert the pre-shock segment into the first scalogram by performing a Morlet wavelet transform on the pre-shock segment, and
wherein the processor is configured to convert the at least one post-shock segment into the at least one second scalogram by performing the Morlet wavelet transform on the at least one post-shock segment.

### Claim 3 (depends on 1)

The medical device of claim 1, wherein the patient is receiving first chest compressions during the time interval preceding administration of the initial electrical shock to the heart of the patient and the processor is further configured to remove artifacts associated with the first chest compressions from the pre-shock segment of the ECG, and/or
wherein the patient is receiving second chest compressions during the time interval following the administration of the initial electrical shock to the heart of the patient and the processor is further configured to reduce artifacts associated with the second chest compressions from the at least one post-shock segment of the ECG.

### Claim 4 (depends on 3)

The medical device of claim 3, wherein the processor is further configured to:
reduce artifacts associated with the second chest compressions from the at least one post-shock segment of the ECG by normalizing wavelet coefficient magnitude values of the at least one second scalogram, and
normalize the first scalograms and the second scalograms by dividing wavelet coefficient magnitude values within each scalogram frequency by at least one of a maximum, median, or mean wavelet coefficient magnitude within each scalogram frequency.

### Claim 5 (depends on 1)

The medical device of claim 1, wherein the treatment comprises administration of an antiarrhythmic medication to the patient, administration of a subsequent electrical shock at a different vector than the initial electrical shock, administration of a double-sequential defibrillation (DSD) treatment, administration of cardiac catheterization treatment, withholding of vasopressor treatment, or administration of extracorporeal membrane oxygenation treatment.

### Claim 6 (depends on 1)

The medical device of claim 1, wherein generating comparisons between the first scalogram and the at least one second scalogram to eigenscalograms comprises:
identifying pre-shock dot products by comparing the pre-shock scalogram to the eigenscalograms;
identifying post-shock dot products by comparing the at least one post-shock scalogram to the eigenscalograms, and
wherein determining that the comparisons are indicative of shock-refractory VF comprises
inputting the pre-shock dot products and the post-shock dot products into a trained machine learning (ML) model; and
receiving a classification that the patient is susceptible to shock-refractory VF from the trained ML model.

### Claim 7 (depends on 1)

The medical device of claim 1, wherein the output device is configured to output the instruction to administer the treatment during a CPR period following the administration of the initial electrical shock to the patient, and
wherein the CPR period ends about 2 minutes after administration of the initial electrical shock to the patient.

### Claim 8

A method, comprising:
identifying a first segment of an electrocardiogram (ECG) that ends prior to administration of an electrical shock;
identifying at least one second segment of the ECG that begins after the administration of the electrical shock;
predicting, based on the first segment and the at least one second segment, that the ECG is indicative of a defibrillation-resistant heart rhythm by:
converting the first segment into a first scalogram;
comparing the first scalogram to eigenscalograms;
converting the at least one second segment into at least one second scalogram; and
comparing the at least one second scalogram to eigenscalograms; and
outputting an indication of the predicted defibrillation-resistant heart rhythm.

### Claim 9 (depends on 8)

The method of claim 8, wherein converting the first segment into a first scalogram comprises performing a wavelet transform on the first segment, and
wherein converting the at least one second segment into at least one second scalogram comprises performing a wavelet transform on the at least one second segment.

### Claim 10 (depends on 8)

The method of claim 8, wherein the at least one second segment comprises chest compression artifacts, and
wherein the method further comprises:
reducing the chest compression artifacts from the at least one segment by normalizing wavelet coefficient magnitude values of the second scalograms.

### Claim 11 (depends on 10)

The method of claim 10, wherein normalizing the wavelet coefficient magnitude values of the second scalograms comprises dividing the wavelet coefficient magnitude values within each scalogram frequency by at least one of a maximum, a median, or a mean wavelet coefficient magnitude within each scalogram frequency.

### Claim 12 (depends on 8)

The method of claim 8, further comprising:
trimming the first scalogram to a first frequency range,
trimming the second scalogram to a second frequency range, and
wherein the first frequency range is broader than the second frequency range.

### Claim 13 (depends on 8)

The method of claim 8, wherein comparing the first scalogram to eigenscalograms comprises determining first dot products of the first segment,
wherein comparing the at least one second scalogram to the eigenscalograms comprises determining second dot products of the at least one second segment, and
wherein determining that the ECG is indicative of the defibrillation-resistant heart rhythm comprises inputting the first dot products and the second dot products into a classifier, the classifier being at least one of a trained random forest model, SVM, or k-nearest neighbor model.

### Claim 14 (depends on 8)

The method of claim 8, the electrical shock being a first electrical shock, the method further comprising:
outputting an instruction to administer an additional treatment,
wherein the additional treatment comprises at least one of administration of an antiarrhythmic medication, administration of a second electrical shock at a different vector than the first electric shock, administration of the second electrical shock at a higher energy level than the first electrical shock, administration of extracorporeal membrane oxygenation, administration of a cardiac catheterization treatment, or administration of DSD.

### Claim 15 (depends on 8)

The method of claim 8, wherein predicting, based on the first segment and the second segment, that the ECG is indicative of the defibrillation-resistant heart rhythm further comprises:
determining that the first segment exhibits VF; and
determining that at least one of a frequency, energy, or magnitude of the VF is below a threshold.

### Claim 16 (depends on 8)

The method of claim 8, wherein outputting the indication of the predicted defibrillation-resistant heart rhythm occurs during a CPR period following the electrical shock, and/or
wherein outputting the indication of the predicted defibrillation-resistant heart rhythm occurs within two minutes of administration of the electrical shock.

### Claim 17 (depends on 8)

The method of claim 8, wherein identifying the at least one second segment of the ECG that begins after the administration of the electrical shock comprises:
identifying a third segment of the ECG that begins after the administration of the electrical shock;
determining that the defibrillation-resistant heart rhythm cannot be predicted based on the first segment and the third segment; and
in response to determining that the defibrillation-resistant heart rhythm cannot be predicted based on the first segment and the third segment, identifying the at least one second segment of the ECG that comprises the third segment and a fourth segment of the ECG that begins after the third segment.

### Claim 18 (depends on 8)

The method of claim 8, wherein the at least one second segment of the ECG comprises multiple segments of the ECG that begin after the administration of the electrical shock, and
wherein the multiple segments occur within about 1 minute after the administration of the electrical shock.

### Claim 19

A computing device, comprising:
at least one processor; and
memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:
identifying a first segment of an electrocardiogram (ECG) that ends prior to administration of an electrical shock;
identifying at least one second segment of the ECG that begins after the administration of the electrical shock;
predicting, based on the first segment and the at least one second segment, that the ECG is indicative of a defibrillation-resistant heart rhythm by:
converting the first segment into a first scalogram;
comparing the first scalogram to eigenscalograms:
converting the at least one second segment into at least one second scalogram; and
comparing the at least one second scalogram to eigenscalograms; and
outputting an indication of the predicted defibrillation-resistant heart rhythm.

### Claim 20 (depends on 19)

The computing device of claim 19, wherein the operations further comprise:
trimming the first scalogram to a first frequency range,
trimming the second scalogram to a second frequency range, and
wherein the first frequency range is broader than the second frequency range.
