Respiratory stress detection

• US 9,750,463 B2
• Filed: 12/10/2013
• Issued: 09/05/2017
• Est. Priority Date: 12/10/2013
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Swamy, Gokul, Genc, Sahika, Ravishankar, Hariharan, Saha, Aditya
• Primary Examiner(s): Kahelin, Michael
• Assistant Examiner(s): Alter, Mitchell E
• Application Number:
  US14/101,663
• Publication Number:
  US 20150157275A1
• Time in Patent Office: 1,365 Days
• CPC Class Codes: A61B 5/002 Monitoring the patient using a local or closed circuit, e.g. in a room or building A61B5/0017 takes precedence
• CPC Class Codes: A61B 5/0022 Monitoring a patient using a global network, e.g. telephone networks, internet
• CPC Class Codes: A61B 5/01 Measuring temperature of body parts ; Diagnostic temperature sensing, e.g. for malignant or inflamed tissue clinical contact thermometers G01K13/20
• CPC Class Codes: A61B 5/02055 Simultaneously evaluating both cardiovascular condition and temperature
• CPC Class Codes: A61B 5/021 Measuring pressure in heart or blood vessels A61B5/0205 takes precedence
• CPC Class Codes: A61B 5/024 Detecting, measuring or recording pulse rate or heart rate A61B5/0205, A61B5/021 take precedence
• CPC Class Codes: A61B 5/0816 Measuring devices for examining respiratory frequency measuring frequency of electric signals G01R23/00
• CPC Class Codes: A61B 5/0836 Measuring rate of CO2 production
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/14542 for measuring blood gases A61B5/14551 takes precedence
• CPC Class Codes: A61B 5/201 Assessing renal or kidney functions
• CPC Class Codes: A61B 5/202 Assessing bladder functions, e.g. incontinence assessment
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/7246 using correlation, e.g. template matching or determination of similarity
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: A61B 5/746 Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms alarms per se G08B
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment

Abstract

Embodiments of the disclosure are directed to a system for analysis of respiratory distress in hospitalized patients. The system performs multi-parametric simultaneous analysis of respiration rate (RR) and pulse oximetry (SpO₂) data trends in order to gauge patterns of patient instability pertaining to respiratory distress. Three patterns in SpO₂ and RR are used along with LOWESS algorithm and Chauvenets criteria for outlier rejection to obtain robust short term and long term trends in RR and SpO₂. Pattern analysis detects the presence of any one of three pattern types proposed. Further, a learning paradigm is introduced to find unknown instances of respiratory distress. This algorithm in conjunction with the learning model allows early detection of respiratory distress in hospital ward and ICU patients.

Claims

1. A computer-implemented method of detecting respiratory distress in a patient, the computer including a processor, said method comprising acts of:

learning, by the processor, a first pattern including a trending classification of respiratory rate and SpO2 using patient history data;

monitoring, by the processor, patient data over a time period, the patient data comprising measures of respiratory rate and SpO2 recorded simultaneously in a storage component;

tracking, by the processor, the measures of respiratory rate and SpO2 over said time period, individually, in corresponding least squares regression models, wherein tracking comprises collecting and storing the measures of respiratory rate and SpO2 in the storage component;

analyzing, by the processor, the least squares regression models of each of the measures of respiratory rate and SpO2 to identify multiple segmented trends in each of the least squares regression models;

removing, by the processor, a noisy deviation from the measures of respiratory rate and SpO2 by using the multiple segmented trends;

identifying, by the processor, the multiple segmented trends in each of the least squares regression models as one of an uptrend, downtrend, or neutral;

determining, by the processor, a second pattern based on the multiple segmented trends from the measures of respiratory rate and the multiple segmented trends from the measures of SpO2;

predicting, by the processor, a potential patient distress by correlating the first pattern to the second pattern; and

triggering, by the processor, an alarm based on the correlation of the first pattern and the second pattern, wherein the alarm is a warning system of patient distress and prevents non-actionable alarms.

2. The computer-implemented method of claim 1, further comprising a step of correlating said measures of respiratory rate and SpO2.

3. The computer-implemented method of claim 2, wherein the first pattern is one of three types comprising:

(a) Type I;

gradual decrease in SpO2 with compensatory hyperventilation;

(b) Type II;

progressive minute decreases in minute ventilation and SpO₂;

or

(c) Type III;

Guarded rapid airflow including SpO₂ reductions followed by precipitous SpO₂ fall.

4. The computer-implemented method of claim 1, wherein the first pattern is learned and unique to said patient.

5. The computer-implemented method of claim 1, wherein said early warning system of patient distress indicates early detection of at least one of cardiopulmonary arrest, respiratory failure, renal failure, sepsis, and re-intubation risk.

6. The computer-implemented method of claim 1, wherein said patient data comprises measures of a plurality of vital signs, in combination.

7. The computer-implemented method of claim 6, wherein said step of tracking includes said measures of said plurality of vital signs.

8. The computer-implemented method of claim 1, wherein the first pattern comprises one of:

(a) hyperventilation compensated respiratory di stress, (b) progressive unidirectional hypoventilation, or (c) sentinel rapid airflow with SpO2 reductions.

9. The computer-implemented method of claim 1, wherein said measures of respiratory rate and SpO2, individually, do not trigger an alarm, and wherein changes in said measures of respiratory rate and SpO2, in combination, trigger an alarm.

10. The computer-implemented method of claim 1, wherein said multiple segmented trends are estimated to predict patterns of progressing patient instability.

11. The computer-implemented method of claim 1, wherein said multiple segmented trends are represented by a 3-tuple time series T:

{TY, t, s} wherein TY denotes each segmented trend as uptrend, downtrend, or neutral;

wherein t is the time duration for which the trend component is active; and

wherein s denotes the strength as a magnitude of mild, moderate, or severe such that the trend is characterized as a label vector T_RR/SpO2=[T₁, T₂ T₃ . . . T_n].

12. The computer-implemented method of claim 11, wherein the label vector is continuously generated and is continuously updated.

13. A computerized system for early detection of respiratory distress comprising:

one or more sensors attached to a patient to monitor a plurality of vital signs;

a monitoring system connected to said one or more sensors;

a storage component connected to said monitoring system to record patient data, wherein patient data comprises measures of the plurality of vital signs including respiratory rate and SpO₂ of the patient; and

a processor interconnected with said monitoring system and said storage component, wherein said processor is configured to;

learn a first pattern including a trending classification of respiratory rate and SpO₂ using patient history data;

analyze the patient data over a time period;

analyze least squares regression models of each of the measures of respiratory rate and SpO₂ to identify segmented trends in the measures of respiratory rate and SpO2 simultaneously;

remove a noisy deviation from the measures of respiratory rate and SpO₂ by using the segmented trends;

determine a second pattern based on the segmented trends in the measures of respiratory rate and SpO₂;

predict a potential patient distress by correlating the first pattern to the second pattern; and

trigger an alarm based on the correlation of the first pattern and the second pattern, wherein the alarm is a warning system of patient distress and prevents non-actionable alarms.

14. The computerized system of claim 13, wherein said segmented trends are identified as one of an uptrend, downtrend, or neutral using a least squares regression model.

15. The computerized system of claim 13, wherein each of said segmented trends in sequential combination establish said second pattern as the patient data is being recorded.

16. The computerized system of claim 13, wherein said measures of respiratory rate and SpO2 are recorded and correlated simultaneously.

17. The computerized system of claim 13, wherein said plurality of vital signs further comprise body temperature, pulse or heart rate, blood pressure, blood glucose, urine production, urinary incontinence, and end-tidal CO₂, which are correlated with said respiratory rate and SpO2, alone or in combination, to characterize said second pattern that alerts a clinician as to patient distress.

18. The computerized system of claim 13, further comprising patient history data integrated with said storage component to characterize said first pattern.

19. The computer-implemented system of claim 13, wherein said second pattern integrates patient use of any medications, sedatives, analgesics, vitamins, and supplements.

20. The computer-implemented system of claim 13, wherein the processor is a computing chip connected wirelessly to a secure network.

21. The computer-implemented system of claim 20, wherein a plurality of web-based personal devices can access said patient data and said patient status.