Noise-robust sleep apnea diagnostic system and related method
Methods and systems for obstructive sleep apnea diagnosis and prediction are disclosed. The methods and systems include: obtaining a white noise contaminated sensor signal for a patient; extracting a feature based on the white noise contaminated sensor signal; determining a matrix based on the feature; determining an intermittent forcing signal based on the matrix; determining an overcomplete representation of the intermittent forcing signal; and generating an obstructive sleep apnea indication based on the overcomplete representation and a threshold. Other aspects, embodiments, and features are also claimed and described.
1 . A method for obstructive sleep apnea detection comprising:
obtaining a white noise contaminated sensor signal for a patient;
extracting a feature based on the white noise contaminated sensor signal;
determining a matrix based on the feature;
determining an intermittent forcing signal based on the matrix;
determining an overcomplete representation of the intermittent forcing signal; and
generating an obstructive sleep apnea indication based on the overcomplete representation and a threshold.
2 . The method of claim 1 , wherein the white noise contaminated sensor data comprises an electrocardiogram signal.
3 . The method of claim 1 , wherein the obtaining of the white noise contaminated sensor signal comprises:
receiving a sensor signal; and
adding white noise in the sensor signal for the white noise contaminated sensor signal.
4 . The method of claim 1 , further comprising:
determining a level of the white noise to make the intermittent forcing signal a Gaussian distribution.
5 . The method of claim 1 , wherein the feature comprises a heart rate variability feature to quantify time intervals between adjacent heartbeats.
6 . The method of claim 1 , wherein the matrix comprises a window length of the feature and a number of points in a trajectory of the feature.
7 . The method of claim 1 , further comprising:
determining a burst duration and an inter-burst duration between two adjacent burst durations; and
determining an obstructive sleep apnea characteristic for the patient based on the burst duration and the inter-burst duration.
8 . The method of claim 7 , wherein the two adjacent burst durations comprise a first burst duration and a second burst duration, and
wherein the inter-burst duration comprises a duration between an end time of the first burst duration and a starting time of the second burst duration.
9 . The method of claim 1 , wherein the overcomplete representation is determined by applying a continuous wavelet transform to the intermittent forcing signal.
10 . The method of claim 9 , wherein the continuous wavelet transform is defined as:
X
w
(
a
,
b
)
=
1
❘
"\[LeftBracketingBar]"
a
❘
"\[RightBracketingBar]"
1
/
2
∫
-
∞
∞
v
r
(
t
)
ψ
¯
(
t
-
b
a
)
dt
,
where X w is the continuous wavelet transform, ψ(t) is a continuous mother wavelet function and ψ is a complex conjugate, a is a scale, b is a translational value, and v r (t) is the intermittent forcing signal.
11 . The method of claim 10 , wherein the threshold is defined as:
|X w (a,b)|≥ψ CWT max|X w (a,b)|, where ψ CWT is a tuning parameter.
12 . The method of claim 1 , wherein the obstructive sleep apnea indication comprises a binary indication indicative of normal breathing or disordered breathing.
13 . The method of claim 1 , further comprising:
receiving a sensor signal;
determining multi-scale system dynamics; and
adaptively sampling the sensor signal based on the multi-scale system dynamics for the white noise contaminated sensor signal.
14 . A system for obstructive sleep apnea detection comprising:
a memory; and
a processor communicatively coupled to the memory;
wherein the memory stores a set of instructions which, when executed by the processor, causes the processor to:
obtain a white noise contaminated sensor signal for a patient;
extract a feature based on the white noise contaminated sensor signal;
determine a matrix based on the feature;
determine an intermittent forcing signal based on the matrix;
determine an overcomplete representation of the intermittent forcing signal; and
generate an obstructive sleep apnea indication based on the overcomplete representation and a threshold.
15 . The system of claim 14 , wherein the white noise contaminated sensor data comprises an electrocardiogram signal.
16 . The system of claim 14 , wherein the feature comprises a heart rate variability feature to quantify time intervals between adjacent heartbeats.
17 . The system of claim 14 , wherein the overcomplete representation is determined by applying a continuous wavelet transform to the intermittent forcing signal.
18 . The system of claim 17 , wherein the continuous wavelet transform is defined as:
X
w
(
a
,
b
)
=
1
❘
"\[LeftBracketingBar]"
a
❘
"\[RightBracketingBar]"
1
/
2
∫
-
∞
∞
v
r
(
t
)
ψ
¯
(
t
-
b
a
)
dt
,
where X w is the continuous wavelet transform, ψ(t) is a continuous mother wavelet function and ψ is a complex conjugate, a is a scale, b is a translational value, and v r (t) is the intermittent forcing signal.
19 . The system of claim 18 , wherein the threshold is defined as:
|X w (a,b)|≥ψ CWT max|X w (a,b)|, where ψ CWR is a tuning parameter.
20 . The system of claim 14 , wherein the set of instructions further causes the processor to:
receive a sensor signal;
determine multi-scale system dynamics; and
adaptively sample the sensor signal based on the multi-scale system dynamics for the white noise contaminated sensor signal.