Sensor Fusion for Physiological Pathology Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing screening methods for physiological pathology, such as COVID-19, face challenges with inaccurate non-contact sensors, long data acquisition times, and inability to detect symptoms that vary between individuals, leading to inefficiencies and inaccuracies in rapid screening processes.
Innovation Solution
A system utilizing multiple remote non-contact sensors of different modalities simultaneously captures datasets to compute a physiological parameter indicative of pathology, combining sub-parameters for enhanced accuracy and reducing data acquisition time, while excluding erroneous data and adjusting sensor orientations for improved measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple remote non-contact sensors of different modalities are used to measure physiological parameters, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple remote non-contact sensors of different modalities (thermal camera, radar, visual camera, acoustic sensor) into a unified sensing system. These sensors simultaneously capture different physiological parameters (temperature, respiratory rate, heart rate, oxygen saturation) and the system merges their data through sensor fusion algorithms to compute an overall physiological status indicator, thereby improving measurement precision while managing device complexity through integrated processing.
Solution Approach 2:
The system employs multi-functional sensors that can detect multiple physiological parameters simultaneously. For example, the thermal camera not only measures body temperature but also detects respiratory patterns through thermal variations, while the radar sensor captures both chest motion for respiratory rate and blood flow patterns for oxygen saturation. This multi-functionality reduces the need for separate dedicated sensors for each parameter.
2Reliability
If data from multiple sensors are combined through sensor fusion, then reliability of physiological parameter detection is improved, but computing requirements and processing time increase
Solution Approach 1:
The system performs preliminary processing of sensor data by extracting relevant physiological features directly from raw sensor signals before fusion. For instance, respiratory rate is extracted from chest motion patterns in radar data, and body temperature is extracted from thermal image sequences prior to combining with other sensor inputs. This preliminary extraction reduces the dimensionality and complexity of data requiring fusion processing.
Solution Approach 2:
The patent segments the physiological parameter detection process into independent modules, each handling specific sensor types and extracting specific physiological features. The thermal processing module, radar processing module, and acoustic processing module operate semi-independently, allowing parallel processing of different sensor streams. This segmentation enables efficient computation while maintaining the reliability benefits of multi-sensor fusion.
3Productivity
If non-contact sensors are used for rapid screening, then productivity is improved, but measurement precision deteriorates due to sensor inaccuracies
Solution Approach 1:
The system merges data from multiple independent non-contact sensors to compensate for individual sensor inaccuracies. By combining thermal, radar, visual, and acoustic sensor measurements through sensor fusion, the system achieves measurement precision that exceeds what any single non-contact sensor could provide, while maintaining rapid screening capability.
Solution Approach 2:
The system uses feedback from multiple sensor modalities to validate and correct measurements in real-time. For example, if thermal sensor readings are ambiguous, the system cross-validates with radar chest motion data and acoustic breathing sounds to confirm respiratory status. This multi-modal feedback mechanism improves measurement precision without significantly increasing processing time.
Data Source
AI summary
There is provided a system for measuring a physiological parameter of a person indicative of physiological pathology, comprising: a plurality of remote non-contact sensors, each of a different type of sensing modality, at least one hardware processor executing a code for: simultaneously receiving over a time interval, from each of the plurality of remote non-contact sensors monitoring a person, a respective dataset, extracting, from each respective dataset, a respective sub-physiological parameter of a plurality of sub-physiological parameters, analyzing a combination of the plurality of sub-physiological parameters, and computing a physiological parameter indicative of physiological pathology according to the analysis, wherein an accuracy of the physiological parameter computed from the combination is higher than an accuracy of the physiological parameter independently computed using any one of the plurality of sub-physiological parameters.