Non-invasive determination of disease states
An example method for estimating a disease state for a patient can include: capturing physiological data including images from the patient over time; processing the physiological data to detect a variability; comparing the variability to a baseline or a range; and estimating the disease state for the patient based upon the comparing of the variability to the baseline or the limit.
1. A method for estimating a disease state for a patient, the method comprising:
capturing images of an unobstructed portion of the patient;
processing the images to detect a pulse rate;
determining a pulse rate variability based upon the pulse rate; and
calculating an early warning score using the pulse rate variability, including:
comparing the pulse rate variability to a baseline having similar demographics; and
increasing or decreasing the early warning score based upon a variance from the baseline.
2. The method of claim 1 , further comprising transmitting pulse rate variability to an electronic medical record.
3. The method of claim 2 , further comprising annotating data with at least one patient outcome metric.
4. The method of claim 3 , further comprising using one or more annotations to measure algorithm performance.
5. The method of claim 1 , wherein estimating the disease state further comprises analyzing patient health factors.
6. The method of claim 1 , further comprising processing the images to detect a respiratory rate for the patient.
7. The method of claim 1 , further comprising forming an autoregressive model using data from the images to estimate the pulse rate of the patient.
8. The method of claim 7 , further comprising combining the autoregressive model estimate of the pulse rate and pulse-rate estimates derived from analysis of motion data.
9. The method of claim 1 , further comprising providing a trend of the pulse rate variability for the patient over time.