CA3094908A1

CA3094908A1 - Smart wearable health monitor system

Info

Publication number: CA3094908A1
Application number: CA3094908A
Authority: CA
Inventors: Shiwei Liu
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-09-30
Filing date: 2020-09-30
Publication date: 2022-03-30

Abstract

A smart wearable health monitoring system is disclosed that creates a medical model of a target individual in a species based on a laser doppler blood vessel imaging scan and a plurality of sensors.

Description

SMART WEARABLE HEALTH MONITOR SYSTEM
TECHNICAL FIELD OF THE INVENTION
[0001] The utility model relates to a smart wearable health monitoring device, belonging to the field of Internet of Things technologies.
Date Recue/Date Received 2020-09-30 BRIEF DESCRIPTION OF THE DRAWINGS

[0002] The present invention is illustrated and described herein with reference to the various drawings, in which like reference numbers denote the various components and/or steps, as appropriate. The drawings illustrate example embodiments of the present disclosure and cannot be considered as limiting its scope.

[0003] FIG. 1 is a schematic diagram of the utility model, according to some example embodiments.

[0004] FIG. 2 is a schematic diagram of the power module, according to some example embodiments.

[0005] FIG. 3 is a flowchart of the machine learning model, according to some example embodiments.
Date Recue/Date Received 2020-09-30 DETAILED DESCRIPTION

[0006] The measurement of vital signs is critical to the treatment and management of many medical conditions for a variety of target individuals or populations. A
smart wearable health monitoring system capable of accurately and frequently measuring vital signs including body temperature, blood pressure, peripheral capillary oxygen saturation (Sp02), pulse rate, and respiration rate is described. The invention will utilize machine learning techniques to infer vital health signals more accurately.

[0007] In various embodiments, the invention utilizes a Laser Doppler Blood Vessel Imaging (LDBVI) device to reconstruct an accurate 4-dimensional representation of the micro blood vessels in the targeted area. In combination with a portable photoplethysmography (PPG) device (used to measure pulse oximetry in clinical settings for pulse and oxygen saturation measurements) to gather data from various sensors and a customized deep learning model, the invention will generate a personalized algorithm to better infer vital health signs in real time.

[0008] In various embodiments, the invention will first attach a custom LDBVI
device to a targeted surface area on the measured individual. Additional sensors including a digital blood pressure gauge and respiration sensors will be used at this stage. The data will be stored on a remote computing device to reconstruct a detailed 4D micro blood vessel image and fused with other sensor readings. Using a custom data visualization platform and automated machine learning platform, the invention can quickly analyze the health-related data from the measured individual and generate a personalized baseline profile. By attaching a custom PPG wearable device to the measured individual, the invention can infer the vital health signals in real time through a custom algorithm.

[0009] In various embodiments, the invention will utilize machine learning techniques to obtain more accurate vital health signals. A custom quantized machine learning model will run on the PPG device and process the sensor data in real time to infer measurements including pulse rate, blood pressure, 5p02, and respiration rate.
The invention will utilize a Deep Neural Network (DNN) to generate a correlation function that converts PPG data in real time to identify key features including systolic blood pressure (SBP), diastolic blood pressure (DBP), 5p02, and respiration rate. The Date Recue/Date Received 2020-09-30 invention will utilize a Recurrent Neural Network (RNN) model to extract more detailed vital signs from the 4D micro blood vessel time-series data captured with the LDBVI device to reconstruct patient's hemodynamics system. The invention will use this data to generate a specific baseline vital health signal data for each measured individual.

[0010] In various embodiments, the invention will track fluctuations to the vital health signal data through the PPG device. The invention will utilize machine learning models to infer if variations from the baseline health signal data are significant in both measured individuals and populations.

[0011] Detailed Description (General Embodiment)

[0012] The utility model relates to a smart wearable health monitoring system, comprising: a heart rate and blood oxygen sensor 106, a temperature sensor 107, an accelerometer and piezo sensor 108, a microcontroller 102, a display screen 103, a wireless module 104, a power module 101, and a user device 105, wherein the various sensors feed the data to the microcontroller, then displays the data on the display screen, and sends the data to the user device by means of a wireless module, and the power module is configured to supply power to the microcontroller.

[0013] The power module comprises of: a battery 210, a power management circuit 211, a charging port, and a charging station 212, where the battery is connected to a power management circuit which feeds power to the microcontroller 202, and the battery is charged by a charging port which receives power from an external charging station.
EMBODIMENTS

[0014] Embodiment 1 refers to a smart wearable health monitoring system which relates to measured individuals in a population.

[0015] Embodiment 2 refers to a smart wearable health monitoring system which relates to marine animals.

[0016] Embodiment 3 relates to a smart wearable health monitoring system which relates to a specific marine animal (key target species are: marine mammals, fish (salmon, etc.), lobster).
Date Recue/Date Received 2020-09-30

Claims

What is claimed is:

1. A method comprising:
concurrently recording input data from a plurality of medical sensors that are medically attached to a target individual of a species wherein the plurality of medical sensors comprise a first photoplethysmography device, a laser doppler blood vessel imaging device, a pulse rate sensor, a blood pressure sensor, a peripheral capillary oxygen saturation sensor, and a respiration rate sensor, wherein the recording of the input data lasts for at least 5 minutes;
parsing the input data to determine temporal correlations between input data received from the first photoplethysmography device and input data received from one or more of the laser doppler blood vessel imaging device, the pulse rate sensor, the blood pressure sensor, the peripheral capillary oxygen saturation sensor, and the respiration rate sensor, wherein the temporal correlations correlate input data from the first photoplethysmography device to input data from the pulse rate sensor, the blood pressure sensor, the peripheral capillary oxygen saturation sensor and the respiration rate sensor;
loading the temporal correlations to a second photoplethysmography device;
after the recording, reading live data from the second photoplethysmography device that is attached to the individual; and determining medical data of the individual comprising a pulse rate of the mammal, a blood pressure of the species, a peripheral capillary oxygen saturation of the individual, and a respiration rate of the individual according to temporal correlations between the live data from the second photoplethysmography device and recorded input data; and storing historical records of the medical data of the individual on a memory device.
Date Recue/Date Received 2020-09-30

2.The method of claim 1 further comprising wirelessly transmitting the historical records to a remote computing device.

3.The method of claim 1, further comprising receiving one or more medical threshold values.

4.The method of claim 3, further comprising transmitting an alert to a remote computing device in response to data in the historical records exceeding one or more of the medical threshold values.

5.The method of claim 1, wherein determining temporal correlations comprises constructing a detailed micro blood vessel four dimensional image by fusing data from the photoplethysmography device and the recorded input data, and training a machine learning model on the input data to generate a medical model of the individual.
Date Recue/Date Received 2020-09-30