SYSTEMS AND METHODS FOR DETERMINING RESPIRATION INFORMATION BASED ON PRINCIPAL COMPONENT ANALYSIS
A patient monitoring system may receive a physiological signal such as a photoplethysmograph (PPG) signal. A plurality of respiration morphology signals may be determined from the PPG signal. Principal component analysis may be performed on the respiration morphology signals, resulting in one or more principal components. Respiration information such as respiration rate may be determined at least in part from a principal component that corresponds to a respiration source signal.
1 . A method comprising:
receiving a photoplethysmograph (PPG) signal;
processing, with processing equipment, the PPG signal to generate a plurality of respiration morphology signals;
performing, with the processing equipment, principal component analysis on the plurality of respiration morphology signals to generate one or more principal components;
identifying, with the processing equipment, a principal component of the one or more principal components that corresponds to a respiration source signal; and
determining, with the processing equipment, respiration information based at least in part on the identified principal component.
2 . The method of claim 1 , wherein the plurality of respiration morphology signals comprise one or more of a down signal, a difference in the second derivative signal, and a kurtosis signal.
3 . The method of claim 1 , wherein processing the PPG signal to generate a plurality of respiration morphology signals comprises:
generating a plurality of candidate respiration morphology signals;
calculating one or more confidence metrics associated with the candidate respiration morphology signals; and
selecting the plurality of respiration morphology signals from the candidate respiration morphology signals based on the confidence metrics.
4 . The method of claim 1 , wherein processing the PPG signal to generate a plurality of respiration morphology signals comprises:
generating a plurality of candidate respiration morphology signals;
determining a predicted respiration range; and
selecting the plurality of respiration morphology signals from the candidate respiration morphology signals based on the predicted respiration range.
5 . The method of claim 1 , wherein the method further comprises:
processing, with processing equipment, the PPG signal to generate a second plurality of respiration morphology signals;
performing, with the processing equipment, principal component analysis on the second plurality of respiration morphology signals to generate one or more second principal components;
identifying, with the processing equipment, a second principal component of the one or more second principal components that corresponds to a respiration source signal; and
determining, with the processing equipment, respiration information based at least in part on the identified principal component and the second identified principal component.
6 . The method of claim 1 , wherein determining the respiration information comprises:
generating a combined signal based on the identified principal component and one or more of the plurality of respiration morphology signals; and
determining the respiration information based on the combined signal.
7 . The method of claim 7 , wherein generating the combined signal comprises:
determining a confidence value for each of the identified principal component and the one or more of the plurality of respiration morphology signals; and
generating a combined signal from the identified principal component and the one or more of the plurality of respiration morphology signals based on the confidence values.
8 . A non-transitory computer-readable storage medium for use in determining respiration information for a patient, the computer-readable medium having computer program instructions recorded thereon for:
receiving a photoplethysmograph (PPG) signal;
processing the PPG signal to generate a plurality of respiration morphology signals;
performing principal component analysis on the plurality of respiration morphology signals to generate one or more principal components;
identifying a principal component of the one or more principal components that corresponds to a respiration source signal; and
determining respiration information based at least in part on the identified principal component.
9 . The computer-readable medium of claim 8 , wherein the plurality of respiration morphology signals comprise one or more of a down signal, a difference in the second derivative signal, and a kurtosis signal.
10 . The computer-readable medium of claim 8 , wherein processing the PPG signal to generate a plurality of respiration morphology signals comprises:
generating a plurality of candidate respiration morphology signals;
calculating one or more confidence metrics associated with the candidate respiration morphology signals; and
selecting the plurality of respiration morphology signals from the candidate respiration morphology signals based on the confidence metrics.
11 . The computer-readable medium of claim 8 , wherein processing the PPG signal to generate a plurality of respiration morphology signals comprises:
generating a plurality of candidate respiration morphology signals;
determining a predicted respiration range; and
selecting the plurality of respiration morphology signals from the candidate respiration morphology signals based on the predicted respiration range.
12 . The computer-readable medium of claim 8 , wherein determining the respiration information comprises:
generating a combined signal based on the identified principal component and one or more of the plurality of respiration morphology signals; and
determining the respiration information based on the combined signal.
13 . The computer-readable medium of claim 12 , wherein generating a combined signal comprises:
determining a confidence value for each of the identified principal component and the one or more of the plurality of respiration morphology signals; and
generating a combined signal from the identified principal component and the one or more of the plurality of respiration morphology signals based on the confidence values.
14 . A patient monitoring system comprising processing equipment configured to:
receive a photoplethysmograph (PPG) signal;
process the PPG signal to generate a plurality of respiration morphology signals;
perform principal component analysis on the plurality of respiration morphology signals to generate one or more principal components;
identify a principal component of the one or more principal components that corresponds to a respiration source signal; and
determine respiration information based at least in part on the identified principal component.
15 . The patient monitoring system of claim 14 , wherein the plurality of respiration morphology signals comprise one or more of a down signal, a difference in the second derivative signal, and a kurtosis signal.
16 . The patient monitoring system of claim 14 , wherein the patient monitoring system is configured to:
generate a plurality of candidate respiration morphology signals;
calculate one or more confidence metrics associated with the candidate respiration morphology signals; and
select the plurality of respiration morphology signals from the candidate respiration morphology signals based on the confidence metrics.
17 . The patient monitoring system of claim 14 , wherein the patient monitoring system is configured to:
generate a plurality of candidate respiration morphology signals;
determine a predicted respiration range; and
select the plurality of respiration morphology signals from the candidate respiration morphology signals based on the predicted respiration range.
18 . The patient monitoring system of claim 14 , wherein the patient monitoring system is configured to:
process the PPG signal to generate a second plurality of respiration morphology signals;
perform principal component analysis on the second plurality of respiration morphology signals to generate one or more second principal components;
identify a second principal component of the one or more second principal components that corresponds to a respiration source signal; and
determine respiration information based at least in part on the identified principal component and the second identified principal component.
19 . The patient monitoring system of claim 14 , wherein the patient monitoring system is configured to:
generate a combined signal based on the identified principal component and one or more of the plurality of respiration morphology signals; and
determine the respiration information based on the combined signal.
20 . The patient monitoring system of claim 14 , wherein the patient monitoring system is configured to:
determine a confidence value for each of the identified principal component and the one or more of the plurality of respiration morphology signals;
generate a combined signal from the identified principal component and the one or more of the plurality of respiration morphology signals based on the confidence values; and
determine the respiration information based on the combined signal.