Touchless Respiration Monitoring via Dynamic Bounding Box
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional medical monitoring systems require physical contact with patients, leading to issues like noise interference from patient movement, limited accuracy, and increased costs, especially in video-based monitoring where noise from environmental and depth camera limitations can obscure respiratory data.
Innovation Solution
A non-contact video-based monitoring system that uses a camera system to capture image data, determines a bounding box and sampling region based on respiratory activity, and generates a volume waveform to accurately measure respiration parameters by combining depth data from multiple regions, reducing noise and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed bounding box is used for sampling in video-based monitoring, then the device complexity is reduced, but the measurement precision deteriorates due to excessive noise from patient movement and environmental factors
Solution Approach 1:
The patent implements dynamic bounding box adjustment by tracking the patient's chest or torso region across multiple video frames. Instead of using a fixed bounding box, the system continuously updates the bounding box position and size based on detected patient movement and respiratory motion, allowing the sampling region to adapt to changing patient positions while maintaining focus on the relevant anatomical area
Solution Approach 2:
The patent applies local quality by selectively sampling pixels only within the dynamically adjusted bounding box that encompasses the patient's chest or torso. This localized sampling approach concentrates computational resources on the most relevant region for respiration detection, improving signal quality by excluding background noise from areas where no respiratory motion occurs
2Measurement precision
If depth data from multiple regions is combined to improve respiration measurement, then the measurement precision improves, but the device complexity and processing requirements increase
Solution Approach 1:
The patent segments the sampling region into multiple discrete regions within the bounding box, such as different zones of the chest or torso. By dividing the overall sampling area into smaller segments, the system can process depth data from each region separately and then combine them to form the complete volume waveform, making the complex processing task more manageable and efficient
Solution Approach 2:
The patent merges depth data from multiple sampled regions within the bounding box to generate the final volume waveform. By combining information from various chest and torso regions, the system creates a more comprehensive and accurate representation of respiratory volume changes, improving measurement precision through data integration
3Measurement precision
If a larger field of view is captured to reduce noise, then the measurement precision improves, but the processing time and computational load increase
Solution Approach 1:
The patent extracts only the essential pixels within the dynamically adjusted bounding box for respiration analysis, rather than processing the entire video frame. By isolating and processing only the relevant pixels that contain respiratory information, the system reduces computational load and processing time while maintaining measurement precision
Data Source
AI summary
An example system includes a camera system, memory, and processing circuitry coupled to the camera system and memory. The processing circuitry is configured to obtain image data and determine a visualization mask based on pixels associated with respiration. The processing circuitry is configured to determine a bounding box based on the visualization mask. The processing circuitry is configured to determine a sampling region for the image data based on an average including the bounding box and at least one previously determined bounding box. The processing circuitry is configured to determine at least a portion of a volume waveform based at least in part on the sampling region, determine at least one respiration parameter based on the volume waveform, and output, for display, the at least one respiration parameter.