Sleep Pulse ML Estimation for Baseline Blood Pressure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing physiological monitoring systems struggle to accurately estimate blood pressure using data from wearable sensors, particularly due to noise and fluctuations in heart rate when a user is awake, leading to inaccurate readings.
Innovation Solution
A method and system that utilize a wearable physiological monitor to obtain pulse data during sleep, determine resting heart rate, and apply a machine learning model trained on features including resting heart rate to predict baseline blood pressure, incorporating demographic information and dynamic pulse features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If pulse data is collected during awake periods, then data collection frequency is high, but measurement precision deteriorates due to noise and heart rate fluctuations
Solution Approach 1:
The system performs preliminary action by collecting pulse data during sleep periods before the actual blood pressure estimation is needed. This advance data collection during low-noise periods prepares clean baseline data that can be used for accurate blood pressure prediction during awake periods, resolving the contradiction between data collection frequency and measurement precision.
Solution Approach 2:
The system introduces sleep period pulse data as an intermediary element. Instead of directly measuring blood pressure during awake periods (which is noisy), it uses clean pulse data from sleep periods as an intermediate step to train machine learning models and predict blood pressure, thereby achieving accurate measurements without direct awake-period measurement.
2Measurement precision
If pulse data is collected during sleep, then measurement precision improves due to reduced noise, but data collection time increases
Solution Approach 1:
The system applies partial action by collecting pulse data during only specific portions of sleep periods when quality is highest, rather than requiring complete sleep session data. This partial data collection achieves sufficient measurement precision without unnecessarily extending data collection time, resolving the contradiction between precision and time loss.
3Quantity of substance
If machine learning models are trained on awake-period data, then training data quantity is high, but model reliability deteriorates due to noise and fluctuations
Solution Approach 1:
The system extracts only the high-quality portion of training data by selecting pulse data collected during sleep periods and excluding noisy awake-period data. This extraction of clean data subsets resolves the contradiction by providing sufficient training data volume from reliable sources, achieving both quantity and reliability requirements.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Provides accurate and non-invasive estimation of baseline blood pressure by reducing noise and fluctuations, enhancing the precision of blood pressure estimation through cleaner pulse data obtained during sleep.
Implementation Method 1
a wearable physiological monitor including a photoplethysmography (PPG) sensor
Data Source
AI summary
A method for baseline blood pressure estimation of a user of a wearable physiological monitor. The method comprising identifying a segment of pulse data related to cardiac activity of the user during a portion of a sleep session, wherein the segment of pulse data is obtained by the wearable physiological monitor; determining, from the segment of pulse data, a resting heart rate value of the user during the portion of the sleep session; identifying a machine learning model trained to receive as input one or more features including an input resting heart rate value obtained during a first time period and predict an indicator of blood pressure during a second time period; and providing the resting heart rate value to the machine learning model to obtain an indicator of baseline blood pressure for the user.