Neural Network Ventilator Work of Breathing Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current ventilator technologies face challenges in accurately and noninvasively predicting a patient's physiologic work of breathing and imposed work of breathing, which are crucial for providing appropriate ventilatory support and determining extubation criteria, due to difficulties in measuring pleural and tracheal pressures.

Innovation Solution

A method and apparatus using a neural network that processes noninvasive parameters from airway pressure, flow, and carbon dioxide sensors to estimate patient effort, including work of breathing, power of breathing, and pressure time product, allowing for more accurate ventilator settings and reduced respiratory muscle fatigue.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If invasive measurement methods (esophageal pressure, tracheal pressure) are used to measure work of breathing, then measurement precision is improved, but ease of operation and patient comfort deteriorate due to invasive procedures

Engineering Contradiction:
Improvework of breathing measurement accuracyVSAvoidpatient comfort and procedural simplicity
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent uses an intermediary mathematical model and neural network algorithm that processes easily obtainable ventilator parameters (flow, airway pressure, volume) to indirectly estimate work of breathing. This intermediary computational approach bridges the gap between simple noninvasive measurements and the accuracy previously only achievable through invasive pressure measurements, eliminating the need for esophageal or tracheal catheters while maintaining measurement precision.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical invasive measurement system (pressure catheters, esophageal balloons) with a computational system using neural networks and mathematical modeling. The mechanical intrusion into the patient's body is substituted by electronic signal processing and algorithmic estimation, achieving the same measurement goal without physical invasion.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Ease of operation

If noninvasive parameters are used to estimate work of breathing, then ease of operation is improved, but measurement precision deteriorates due to inability to directly measure pleural and tracheal pressures

Engineering Contradiction:
Improvenoninvasive measurement capabilityVSAvoidwork of breathing estimation accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent transforms the relationship between measurable parameters and work of breathing by using neural network algorithms that learn complex nonlinear relationships from training data. Instead of relying on simple linear assumptions, the system changes the parameter transformation approach from direct calculation to adaptive computational modeling, significantly improving estimation accuracy from noninvasive inputs.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system incorporates feedback mechanisms where the neural network is trained using actual invasive measurements as ground truth, then uses the learned model to continuously estimate work of breathing from noninvasive parameters. The feedback loop between training data and operational estimation refines the algorithm's accuracy over time, closing the precision gap between invasive and noninvasive methods.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If complex invasive monitoring is implemented to accurately assess patient effort, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvepatient effort measurement accuracyVSAvoidmonitoring system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent makes the ventilator's existing sensors (flow, pressure, volume) perform multiple functions: they continue to provide standard ventilatory monitoring while simultaneously serving as inputs for work of breathing estimation through the neural network. This multi-functionality eliminates the need for separate invasive monitoring equipment, reducing overall device complexity while maintaining measurement precision.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent extracts the work of breathing estimation function from the complex invasive monitoring system and implements it as a separate computational module using existing ventilator data. By taking out this specific function and implementing it algorithmically, the system avoids the complexity of physical invasive sensors while achieving the same measurement capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2816952B1Method and apparatus for predicting work of breathing
Publication Date: 2022.08.03 UNIV OF FLORIDA RESEARCH FOUNDATION INC
  • EP2816952B1 patent drawingFigure 1
  • EP2816952B1 patent drawingFigure 2
  • EP2816952B1 patent drawingFigure 3

AI summary

A method of creating a noninvasive predictor of both physiologic and imposed patient effort of breathing from airway pressure and flow sensors attached to the patient using an adaptive mathematical model. The patient effort is commonly measured via work of breathing, power of breathing, or pressure-time product of esophageal pressure and is important for properly adjusting ventilatory support for spontaneously breathing patients. The method of calculating this noninvasive predictor is based on linear or non-linear calculations using multiple parameters derived from the above-mentioned sensors.