Wearable Biosensor Potassium Monitoring via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wearable biosensing devices are limited in monitoring health characteristics like blood flow beyond the wrist area and provide superficial metrics that can be misleading, failing to indicate vital signs such as pulsatile vascular blood flow, which is crucial for managing conditions like hyperkalemia in patients with kidney disease.
Innovation Solution
A closed-loop architecture that integrates data from wearable biosensing devices with peripheral devices and patient demographic data, using machine learning models to generate treatment recommendations and risk stratifications, enabling continuous monitoring of electrolyte levels and vascular health without blood draws.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of moving object
If wearable biosensing devices are used to monitor health metrics, then continuous monitoring capability is improved, but measurement precision of vital signs like pulsatile vascular blood flow deteriorates
Solution Approach 1:
The patent uses machine learning models as an intermediary to bridge the gap between wearable device data and clinical-grade measurements. The system processes raw biosensing data through trained ML algorithms that have been taught to recognize patterns corresponding to pulsatile vascular blood flow, enabling accurate measurement without direct clinical measurement equipment.
Solution Approach 2:
The patent replaces traditional mechanical/physical measurement methods with computational approaches. Instead of using complex hardware to directly measure pulsatile vascular blood flow, the system substitutes computational analysis of wearable device data with machine learning models that can infer these measurements from available sensor data.
2Measurement precision
If traditional blood draw methods are used for electrolyte monitoring, then measurement precision of potassium levels is improved, but loss of time and patient discomfort increase
Solution Approach 1:
The patent creates a virtual copy of the blood test process through machine learning. Instead of repeatedly performing physical blood draws, the system uses ML models to generate continuous estimates of potassium levels from wearable device data, effectively copying the information-gathering function without the invasive procedure.
Solution Approach 2:
The system enables patients to continuously monitor their own electrolyte levels through wearable devices and ML processing, eliminating the need for frequent clinic visits and blood draws. The machine learning model continuously processes data and provides ongoing assessment without requiring repeated manual intervention.
3Ease of operation
If display-based wearable devices are used, then ease of operation is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts the display and processing functions from the wearable device itself and relocates them to a separate computing system. The wearable device becomes a simple data collection sensor, while the machine learning processing and user interface are handled by external devices like smartphones or computers, reducing the complexity of the wearable component.
Data Source
AI summary
Systems and methods disclosed herein are directed to determining serum potassium levels of a patient wearing a biosensing device. A first such method includes operations of capturing a first energy measurement reading by an energy detecting element of an optical sensor, wherein the optical sensor is a component of a biosensing device, and wherein the biosensing device is disposed on a skin surface of the patient, performing a feature extraction on the first energy measurement reading resulting in a feature vector representative of volumetric variations in blood flow of the patient, deploying a trained machine learning model configured to take the feature vector as input and determine a serum potassium level classification of the patient at a time corresponding to when the first energy measurement reading was captured, and generating a graphic user interface (GUI) that displays the serum potassium level classification.