Method for predicting apnea-hypopnea index from overnight pulse oximetry readings

• US 20050131283A1
• Filed: 09/23/2004
• Published: 06/16/2005
• Est. Priority Date: 09/23/2003
• Status: Active Grant

• Current Assignee: US Department of Veterans Affairs (Government of the United States of America)
• Original Assignee: US Department of Veterans Affairs (Government of the United States of America)
• Inventors: Grant, Brydon J.B., Dmochowski, Jacek
• Granted Patent: US 7,309,314 B2
• Time in Patent Office: Days
• US Class Current: 600/323
• CPC Class Codes: A61B 5/00 Measuring for diagnostic purposes radiation diagnosis A61B6/00; diagnosis by ultrasonic, sonic or infrasonic waves A61B8/00; Identification of persons
• CPC Class Codes: A61B 5/08 Detecting, measuring or recording devices for evaluating the respiratory organs A61B5/0205 takes precedence
• CPC Class Codes: A61B 5/4818 Sleep apnoea
• 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

Abstract

The present invention provides a predictive tool for predicting Apopnea Hypopnea Index (AHI) in the diagnosis of Obstructive sleep apnea (OSA). The predictive tool is developed by recording pulse oximetry readings, obtaining delta index, oxygen saturation times and oxygen desaturation events from the oximetry readings. A multivariate non-parametric analysis and bootstrap aggregation is performed to obtain predictive models which can be used to predictive AHI in an individual and to classify an individual as having or not having OSA.

Claims

1. A method for developing a predictive tool for predicting Apopnea Hypopnea Index (AHI) in an individual comprising the steps of:

a) identifying patients having obstructive sleep apnea;

b) determining actual AHI in the patients;

c) obtaining pulse oximetry recordings from the patients;

d) determining patient data comprising a set of indices from the pulse oximetry readings comprising;

delta index, time spent at selected oxygen saturation levels and oxygen desaturation events;

e) performing random sampling with replacement of patient data from (d) to generate a plurality of data sets;

f) performing multivariate non-parametric analysis in the plurality of data sets to generate a plurality of prediction models;

g) obtaining a provisional predictive AHI value for each patient corresponding to each prediction model; and

h) comparing the provisional predictive AHI value to actual AHI value for each patient to provide weight to each prediction model,

wherein the plurality of prediction models and weight of the prediction models provides a predictive tool for predicting AHI in an individual.

2. The method of claim 1 further comprising the step of validating the prediction models.

3. The method of claim 1, wherein the pulse oximetry readings are obtained overnight.

4. The method of claim 1, wherein the selected oxygen saturation is in the range of 50 to 100%.

5. The method of claim 4, wherein the selected oxygen saturation levels are selected from the group consisting of 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% and combinations thereof.

6. The method of claim 5, wherein the selected oxygen saturation levels are 80, 82, 84, 86, 88 and 90%.

7. The method of claim 1, wherein the oxygen desaturation levels for desaturation events are selected from the group consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20% and combinations thereof.

8. The method of claim 7, wherein the oxygen desaturation levels are 2, 3 and 4%.

9. The method of claim 1 further comprising:

a) obtaining additional indices in claim 1(c) selected from the group consisting of;

electroencephalogram (EEG), electro-oculogram, submental and bilateral leg electromyograms, electrocardiogram, airflow measured by an oral-nasal thermistor, respiratory effort measured by thoracoabdominal piezoelectric belts and responses to sleep history questionnaires; and

b) utilizing the additional indices in the step of claim 1(e).

10. A method for predicting AHI in an individual comprising:

a) obtaining pulse oximetry recordings from the individual;

b) determining a set of indices from the oximetry readings comprising, delta index, time spent at selected oxygen saturation levels and oxygen desaturation events;

c) subjecting the set of indices to a plurality of prediction models of claim 1(f) to obtain a plurality of provisional predicted AHI values corresponding to each model; and

d) obtaining a final predicted AHI value by averaging the weighted provisional predicted AHI values.

11. The method of claim 10, wherein the pulse oximetry readings are obtained overnight.

12. The method of claim 10, wherein the selected oxygen saturation is in the range of 50 to 100%.

13. The method of claim 12, wherein the selected oxygen saturation levels are selected from the group consisting of 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% and combinations thereof.

14. The method of claim 13, wherein the selected oxygen saturation levels are 80, 82, 84, 86, 88 and 90%.

15. The method of claim 10, wherein the oxygen desaturation levels for desaturation events are selected from the group consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20% and combinations thereof.

16. The method of claim 7, wherein the oxygen desaturation levels are 2, 3 and 4%.

17. The method of claim 10, further comprising the step of classifying an individual as having or not having obstructive sleep apnea based on the final predicted AHI value.

18. The method of claim 10, wherein step a) further comprises obtaining additional indices selected from the group consisting of:

electroencephalogram (EEG), electro-oculogram, submental and bilateral leg electromyograms, electrocardiogram, airflow measured by an oral-nasal thermistor, respiratory effort measured by thoracoabdominal piezoelectric belts and responses to sleep history questionnaires; and

step c) further comprises inputting the additional indices into the prediction models to obtain the provisional predicted AHI values.

19. A computational system having software thereon to execute the steps comprising:

a) receiving pulse oximetry readings;

b) calculating indices comprising;

delta index, time spent at selected oxygen saturation levels and oxygen desaturation events;

c) subjecting the indices to a plurality of AHI prediction models to generate provisional AHI values corresponding to each prediction model wherein each prediction model has specific weight assigned thereto; and

d) calculating a final AHI value based on a weighted average of the provisional AHI value.

20. The computational system described in claim 19, wherein:

a) time spent at select oxygen saturation levels is selected from the group consisting of 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% and combinations thereof; and

b) oxygen desaturation levels for desaturation events are selected from the group consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20% and combinations thereof.

21. The computational system of claim 19 further comprising software wherein step a) receives additional indices selected from the group consisting of:

electroencephalogram (EEG), electro-oculogram, submental and bilateral leg electromyograms, electrocardiogram, airflow measured by an oral-nasal thermistor, respiratory effort measured by thoracoabdominal piezoelectric belts and responses to sleep history questionnaires; and

step c) subjects the additional indices to the prediction models.

22. A computational system having software thereon to execute the steps comprising:

a) receiving indices comprising;

delta index, time spent at selected oxygen saturation levels and oxygen desaturation events;

b) subjecting the indices to a plurality of AHI prediction models to generate provisional AHI values corresponding to each prediction model wherein each prediction model has specific weight assigned thereto; and

c) calculating a final AHI value based on a weighted average of the provisional AHI value.

23. The computational system of claim 22, wherein:

a) time spent at select oxygen saturation levels is selected from the group consisting of 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99% and combinations thereof; and

b) oxygen desaturation levels for desaturation events are selected from the group consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20% and combinations thereof.

24. The computational system of claim 22 further comprising software wherein step a) receives additional indices selected from the group consisting of:

electroencephalogram (EEG), electro-oculogram, submental and bilateral leg electromyograms, electrocardiogram, airflow measured by an oral-nasal thermistor, respiratory effort measured by thoracoabdominal piezoelectric belts and responses to sleep history questionnaires; and

step c) subjects the additional indices to the prediction models.