System and methods for video-based monitoring of vital signs
The present invention relates to the field of medical monitoring, and in particular non-contact, video-based monitoring of pulse rate, respiration rate, motion, and oxygen saturation. Systems and methods are described for capturing images of a patient, producing intensity signals from the images, filtering those signals to focus on a physiologic component, and measuring a vital sign from the filtered signals.
1. A video-based method of measuring a patient's vital sign, comprising:
generating a video signal from a video camera having a field of view that includes exposed skin of a patient;
extracting from the video signal a plurality of time-varying intensity signals;
combining two or more of the time-varying intensity signals within a region of the field of view to produce a regional intensity signal;
transforming the regional intensity signal into the frequency domain to produce a regional frequency signal;
accumulating peaks in the regional frequency signal over a sliding time window;
selecting a median frequency from the accumulated peaks;
converting the median frequency into a measured vital sign value and adding the measured vital sign value to a running average; and
outputting the averaged vital sign value for further processing or display.
2. The method of claim 1 , wherein the region is selected based on a strength of modulations of intensity signals in the region.
3. The method of claim 1 , wherein the region includes non-adjacent areas or pixels.
4. The method of claim 1 , further comprising removing peaks from the accumulated peaks when they reach an age limit.
5. The method of claim 1 , further comprising discarding peaks outside of a physiologic limit.
6. The method of claim 1 , further comprising discarding the measured vital sign value when it differs from the running average by more than a defined amount.
7. The method of claim 1 , further comprising discarding peaks if they are sub-harmonics of already identified peaks.
8. The method of claim 1 , further comprising identifying a frequency content of each of said plurality of time-varying intensity signals, averaging the color signals from said plurality of time-varying intensity signals, and de-noising the signal.
9. The method of claim 8 , further comprising performing a fast Fourier transform operation over a sliding time window to identify frequency components of said plurality of time-varying intensity signals.
10. The method of claim 1 , wherein the vital sign comprises respiration rate or heart rate.
11. The method of claim 10 , wherein the vital sign comprises heart rate.
12. A video-based method of measuring a patient's vital sign, comprising:
generating a video signal from a video camera having a field of view that includes exposed skin of a patient;
extracting from the video signal a time-varying color signal for each of a plurality of pixels in the field of view;
identifying a frequency content of each color signal;
combining color signals within a region of the field of view;
measuring said vital sign from the combined color signal, including over a period of time, accumulating frequency peaks from the combined signal, selecting a median frequency and updating a running average vital sign value of a patient by converting the median frequency into a measured vital sign value and adding the measured vital sign value to the running average, wherein accumulating frequency peaks comprises cross-correlating the frequency spectrum of each color signal to amplify results; and
outputting the updated average vital sign value for further processing or display.
13. The method of claim 12 , wherein said colors comprise red, green and blue, and wherein cross-correlating the frequency spectrum of each color signal provides four spectrums that are analyzed to select and accumulate peaks.
14. The method of claim 13 , wherein said cross-correlated spectrum is calculated by multiplying or summing existing spectrums together.
15. The method of claim 14 , wherein individual spectrums are scaled prior to multiplying or summing based on weighting parameters.
16. The method of claim 15 , wherein said weighting parameters are based upon the relative numbers of colors of pixels of said video camera.
17. The method of claim 12 , further comprising de-noising the color signal.
18. The method of claim 12 , wherein said vital sign comprises respiration rate or heart rate.
19. The method of claim 18 , wherein said vital sign comprises heart rate.
20. A video-based method of measuring a patient's vital sign, comprising:
generating a video signal from a video camera having a field of view that includes exposed skin of a patient;
extracting from the video signal a time-varying color signal for each of a plurality of pixels in the field of view;
identifying a frequency content of each color signal;
combining color signals within a region of the field of view;
measuring said vital sign from the combined color signal, including over a period of time, accumulating frequency peaks from the combined signal, selecting a median frequency and updating a running average vital sign value of a patient by converting the median frequency into a measured vital sign value and adding the measured vital sign value to the running average, wherein accumulating frequency peaks comprises adding frequencies multiple times based on a peak's relative height, with harmonics of already added frequencies only added once; and
outputting the updated average vital sign value for further processing or display.
21. The method of claim 20 , wherein said vital sign comprises heart rate.