# US 2026/0225498 A1 · Systems and methods for detecting heartbeat and measuring heart rate

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

- Number: US 2026/0225498 A1
- Document ID: US20260225498A1
- Status: Published application
- Kind: A1
- Application number: 19/042,063
- Filed: 2025-01-31
- Published: 2026-08-06
- Earliest claimed: 2025-01-31
- Assignees: Toyota Motor Engineering & Manufacturing North America, Inc.
- Applicants: Toyota Motor Engineering & Manufacturing North America, Inc.
- Inventors: Frederico Marcolino Quintao Severgnini; Paul Donald Schmalenberg; Bryan Pardo
- Web page: https://patentgrep.com/patent/US20260225498A1

## CPC

- A61B 5/7267: {involving training the classification device}
- B60N 2/0022: {for sensing anthropometric parameters, e.g. heart rate or body temperature}
- A47C 17/86: Parts or details specially adapted for beds, sofas or couches not fully covered by any single one of groups A47C17/02 - A47C17/84
- A61B 5/0245: by using sensing means generating electric signals, {i.e. ECG signals}
- A61B 5/18: for vehicle drivers {or machine operators}
- A61B 5/352: Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
- A61B 5/6893: {Cars}
- A61B 5/7203: {for noise prevention, reduction or removal}
- A61B 5/7207: {of noise induced by motion artifacts}
- A61B 5/7264: {Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems}
- A61B 2503/22: Motor vehicles operators, e.g. drivers, pilots, captains
- B64D 11/06: {Arrangements of seats, or adaptations or details specially adapted for aircraft seats (crew seat constructions for emergency purposes B64D25/04)}
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- B60N: SEATS SPECIALLY ADAPTED FOR VEHICLES; VEHICLE PASSENGER ACCOMMODATION NOT OTHERWISE PROVIDED FOR
- A47C: CHAIRS; SOFAS; BEDS
- B64D: EQUIPMENT FOR FITTING IN OR TO AIRCRAFT; FLIGHT SUITS; PARACHUTES; ARRANGEMENT OR MOUNTING OF POWER PLANTS OR PROPULSION TRANSMISSIONS IN AIRCRAFT

## Abstract

Systems and methods for detecting the heartbeat and measuring the heart rate of a human subject are disclosed herein. In one embodiment, a system receives signals from a plurality of sensors in a seat in which a human subject is sitting. The system denoises the signals using bandpass and wavelet transform filters. The system processes the denoised signals using a machine-learning-based model that performs further denoising and beamforming to detect instances of the heartbeat of the human subject and to measure the heart rate of the human subject. The machine-learning-based model is trained using ground-truth data in which an electrocardiogram waveform is replaced by Gaussian curves centered around R-peaks in the electrocardiogram waveform. The system outputs at least one of (1) one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject.

## Claims

### Claim 1

A system, comprising:
a seat that includes a plurality of sensors;
a processor; and
a memory storing machine-readable instructions that, when executed by the processor, cause the processor to:
receive signals from the plurality of sensors while a human subject is sitting in the seat;
denoise the signals using a bandpass filter and a wavelet transform filter to produce denoised signals;
process the denoised signals using a machine-learning-based model that performs further denoising and beamforming to detect instances of a heartbeat of the human subject and to measure a heart rate of the human subject, wherein the machine-learning-based model is trained using ground-truth data in which an electrocardiogram waveform is replaced by Gaussian curves centered around R-peaks in the electrocardiogram waveform; and
output at least one of (1) one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject.

### Claim 2 (depends on 1)

The system of claim 1, wherein the machine-readable instructions include further instructions that, when executed by the processor, cause the processor to take automatically, in response to the at least one of (1) the one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject, an action to assist the human subject.

### Claim 3 (depends on 2)

The system of claim 2, wherein the seat and the human subject are in a vehicle and the action to assist the human subject includes controlling, at least in part, operation of the vehicle.

### Claim 4 (depends on 1)

The system of claim 1, wherein the seat is one of a vehicle seat, an airplane seat, a medical chair, a medical bed, and a sofa chair.

### Claim 5 (depends on 1)

The system of claim 1, wherein the sensors include one or more of accelerometers, triaxial sensors, and hydrophones.

### Claim 6 (depends on 1)

The system of claim 1, wherein the machine-learning-based model includes a neural network based on a U-Net architecture and an attention block that function together.

### Claim 7 (depends on 1)

The system of claim 1, wherein the machine-learning-based model is trained using a double loss function that combines a Log-Cosh loss function and a sum-of-squared-errors loss function.

### Claim 8

A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to:
receive signals from a plurality of sensors in a seat in which a human subject is sitting;
denoise the signals using a bandpass filter and a wavelet transform filter to produce denoised signals;
process the denoised signals using a machine-learning-based model that performs further denoising and beamforming to detect instances of a heartbeat of the human subject and to measure a heart rate of the human subject, wherein the machine-learning-based model is trained using ground-truth data in which an electrocardiogram waveform is replaced by Gaussian curves centered around R-peaks in the electrocardiogram waveform; and
output at least one of (1) one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject.

### Claim 9 (depends on 8)

The non-transitory computer-readable medium of claim 8, wherein the instructions include further instructions that, when executed by the processor, cause the processor to take automatically, in response to the at least one of (1) the one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject, an action to assist the human subject.

### Claim 10 (depends on 9)

The non-transitory computer-readable medium of claim 9, wherein the seat and the human subject are in a vehicle and the action to assist the human subject includes controlling, at least in part, operation of the vehicle.

### Claim 11 (depends on 8)

The non-transitory computer-readable medium of claim 8, wherein the seat is one of a vehicle seat, an airplane seat, a medical chair, a medical bed, and a sofa chair.

### Claim 12 (depends on 8)

The non-transitory computer-readable medium of claim 8, wherein the machine-learning-based model includes a neural network based on a U-Net architecture and an attention block that function together.

### Claim 13 (depends on 8)

The non-transitory computer-readable medium of claim 8, wherein the machine-learning-based model is trained using a double loss function that combines a Log-Cosh loss function and a sum-of-squared-errors loss function.

### Claim 14

A method, comprising:
receiving signals from a plurality of sensors in a seat in which a human subject is sitting;
denoising the signals using a bandpass filter and a wavelet transform filter to produce denoised signals;
processing the denoised signals using a machine-learning-based model that performs further denoising and beamforming to detect instances of a heartbeat of the human subject and to measure a heart rate of the human subject, wherein the machine-learning-based model is trained using ground-truth data in which an electrocardiogram waveform is replaced by Gaussian curves centered around R-peaks in the electrocardiogram waveform; and
outputting at least one of (1) one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject.

### Claim 15 (depends on 14)

The method of claim 14, further comprising taking automatically, in response to the at least one of (1) the one or more detected instances of the heartbeat of the human subject and (2) the measured heart rate of the human subject, an action to assist the human subject.

### Claim 16 (depends on 15)

The method of claim 15, wherein the seat and the human subject are in a vehicle and the action to assist the human subject includes controlling, at least in part, operation of the vehicle.

### Claim 17 (depends on 14)

The method of claim 14, wherein the seat is one of a vehicle seat, an airplane seat, a medical chair, a medical bed, and a sofa chair.

### Claim 18 (depends on 14)

The method of claim 14, wherein the sensors include one or more of accelerometers, triaxial sensors, and hydrophones.

### Claim 19 (depends on 14)

The method of claim 14, wherein the machine-learning-based model includes a neural network based on a U-Net architecture and an attention block that function together.

### Claim 20 (depends on 14)

The method of claim 14, wherein the machine-learning-based model is trained using a double loss function that combines a Log-Cosh loss function and a sum-of-squared-errors loss function.
