Machine Learning RAP Estimation from IVC Ultrasound Sniff Tests
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating right atrial pressure (RAP) through the sniff test are subjective and prone to operator variability, resulting in inaccurate measurements and increased resource utilization.
Innovation Solution
The development of a machine learning-based system that utilizes video scan data from ultrasound imaging to automatically evaluate RAP, including the identification of sniff tests and estimation of RAP through trained models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual evaluation of sniff test by cardiologists is used, then diagnostic accuracy can be maintained, but operator variability and human error increase
Solution Approach 1:
The patent creates a digital copy of the cardiologist's evaluation process through machine learning models. The system captures ultrasound video frames, automatically identifies IVC anatomy, measures diameter and collapsibility, and predicts RAP without human intervention. This digital copy eliminates operator variability while maintaining diagnostic accuracy through trained algorithms that replicate and extend cardiologist expertise.
Solution Approach 2:
The patent replaces the mechanical human evaluation system with an automated computational system. Instead of cardiologists manually measuring and interpreting ultrasound images, the system uses computer vision algorithms to detect IVC, track its collapse during sniffing, calculate collapsibility percentages, and predict RAP values. This substitution eliminates human error and ensures consistent measurement precision.
2Measurement precision
If automated machine learning evaluation is implemented, then measurement precision and consistency improve, but device complexity increases
Solution Approach 1:
The patent divides the complex RAP estimation task into separate functional modules: (1) IVC anatomy identification, (2) video frame processing, (3) diameter measurement, (4) collapsibility calculation, and (5) RAP prediction. Each module handles a specific aspect of the evaluation, making the overall system more manageable and easier to implement while maintaining high measurement precision through specialized processing at each stage.
Solution Approach 2:
The patent introduces an intermediary processing layer between the ultrasound hardware and the final RAP prediction. This intermediary system includes image processing algorithms that automatically detect IVC, measure dimensions, and prepare data for prediction models. The intermediary layer simplifies the interface between complex imaging hardware and the evaluation logic, making the overall system more manageable while maintaining precision.
3Measurement precision
If right heart catheterization is performed for accurate RAP measurement, then measurement precision is maximized, but invasiveness and resource utilization increase
Solution Approach 1:
The patent creates a non-invasive digital copy of the invasive catheter measurement process. Instead of physically inserting a catheter into the right heart, the system uses machine learning models trained on catheter measurements to predict RAP values from non-invasive ultrasound video data. This copying approach maintains measurement precision while completely eliminating the invasiveness and associated risks of catheterization.
Solution Approach 2:
The patent replaces the mechanical invasive catheterization system with a computational prediction system. Instead of using physical catheters and pressure transducers to directly measure RAP, the system uses computer vision and machine learning algorithms to infer RAP from ultrasound images of IVC collapse. This substitution eliminates all harmful invasive factors while maintaining or improving measurement precision through advanced computational methods.
4Ease of operation
If manual IVC diameter measurement is performed, then ease of operation is maintained, but measurement precision deteriorates due to operator variability
Solution Approach 1:
The patent enables the system to measure IVC diameter automatically without requiring operator skill or intervention. The machine learning model processes ultrasound video frames, identifies the IVC anatomy, tracks its dimensions throughout the sniffing maneuver, and calculates collapsibility percentages. This self-service capability maintains operational simplicity for users while dramatically improving measurement precision through consistent, repeatable automated measurements that do not vary between operators.
Data Source
AI summary
Automated evaluation of RAP via machine learning is described herein. In one implementation, a method includes: obtaining first video scan data including multiple first video frames of an IVC of a subject, the multiple first video frames including a first video frame of the IVC while the subject is at rest, and a second video frame of the IVC while the subject is inhaling; determining, using a first trained model, based at least on the multiple first video frames, that the first video scan data corresponds to a sniff test of the IVC of the subject; and predicting, using a second trained model, based at least on the multiple first video frames, a RAP of the subject.