Robust Estimation of Respiratory Rate using Probabilistic Methods
Measuring the respiratory rate of a patient is traditionally an estimate riddled with mistakes; physicians and automated equipment are both prone to errors. Oxford researchers have invented a probabilistic approach that measures the respiratory rate whilst calculating the confidence of the said measurement. This allows for the first time to measure robustly an important vital sign.
Existing methods of estimating respiratory rate are often manual. The attending clinician counts the number of movements of the chest wall within 15 or 30 seconds. This number is then multiplied by a factor of 4 or 2, respectively, to estimate ‘breaths per minute’. Recent studies have shown this method is unreliable, with the actual respiratory rate often different to that estimated by a clinician. Existing automated methods for estimating respiratory rate from physiological signals (i.e. Electrocardiogram –ECG or photoplethysmography -PPG) are often unreliable when applied to actual patients.
Description
Researchers at the University of Oxford have developed a probilistic approach to measuring the respiratory rate of a patient. The technology gives a probability distribution of likely respiratory rates, rather than a single estimate. This allows the system to produce robust estimates in the presence of noise and artefacts, and where the outputs are useful for patient monitoring systems that take a probabilistic approach.
Technology maturity and readiness
This invention can be applied to all existing respiratory rate monitoring equipment whether these are contact based (ECG, PPG) or contactless (i.e. camera based). The methodology has been validated using signals from forty (40) healthy volunteers, and the results were published in a peer-reviewed conference. The performance of the method matches the performance of the state of the art, whilst bringing the benefits of a probabilistic framework and estimates of confidence.
about this technology
