Biological marker and methods
Systems, methods, and other techniques for monitoring, including non-invasive monitoring, of biological markers based on the interaction, temporal association, or coincidence of brain activity and periphery activity in a mammal are provided. Systems and methods for generating a behavioral state-independent representation of cardiac activity and for identifying cardiac events and/or brain-periphery, e.g., brain-cardiac, temporal associations useful as biomarkers of disease such as, e.g., neurologic disease, in a mammal are also provided.
1. A method for estimating the risk of occurrence of a physiological event in a mammal prior to the occurrence of the physiological event, comprising:
obtaining brain activity data that describes brain activity of the mammal over a first period of time;
obtaining peripheral activity data that describes peripheral activity of the mammal over a second period of time;
receiving, by at least one device, brain data from a brain activity measurement device that describes brain activity of a mammal over the first period of time;
receiving, by the at least one device, periphery data from a peripheral activity measurement device that describes peripheral activity of a mammal over the second period of time;
generating, by the at least one device, one or more brain feature representations based on the brain data;
generating, by the at least one device, one or more periphery feature representations based on the periphery data;
generating, by the at least one device, one or more brain statistical distributions of the one or more brain feature representations;
generating, by the at least one device, one or more periphery statistical distributions of the one or more periphery feature representations;
identifying, by the at least one device, one or more brain feature representations that fall outside of a predetermined range in the brain statistical distribution to generate a set of one or more target brain feature representations;
identifying, by the at least one device, one or more periphery feature representations that fall outside of a predetermined range in the periphery statistical distribution to generate a set of one or more target periphery feature representations;
determining, by the at least one device, a brain-periphery temporal association between the one or more target brain feature representations and the one or more target periphery feature representations;
determining, by the at least one device, an estimate of risk of a future physiological event based on the brain-periphery temporal association; and
providing, by the at least one device, an indication of the estimate of risk of the future physiological event.
2. The method of claim 1 , wherein:
the periphery data is ECG data;
receiving the periphery data comprises receiving, by at least one device, physiological data that describes cardiac activity of a mammal over a period of time;
identifying, by the at least one device, two or more reference points in the physiological data;
determining, by the at least one device, one or more intervals wherein each interval is an interval between a unique set of two of the reference points;
processing, by the at least one device, the one or more intervals to generate one or more cardiac activity feature representations, wherein a distribution of the cardiac activity feature representations of the mammal over the period of time is independent of the mammal's behavioral state during the period of time, wherein processing the one or more intervals comprises determining for each interval a discrete estimate of the logarithmic derivative of the one or more intervals at a selected reference time, proportional to the difference in the interval before and after each selected reference time determined over a first time window, normalized by the average interval over a second time window, and wherein processing the one or more intervals comprises determining dlnRR at any time, T, using first window length τ 1 and second window length τ 2 , according to the equation:
d
ln
RR
(
T
)
τ
1
,
τ
2
⩵
def
1
τ
1
RRI
_
τ
1
(
T
+
τ
1
2
)
-
RRI
_
τ
1
(
T
-
τ
1
2
)
RRI
_
τ
2
(
T
)
wherein
RRI
_
τ
1
(
T
)
=
τ
1
N
RR
,
τ
1
(
T
)
RRI
_
τ
2
(
T
)
=
τ
2
N
RR
,
τ
2
(
T
)
N
RR
,
τ
i
(
T
)
⩵
def
fractional
number
of
intervals
∈
[
T
-
τ
i
2
,
T
+
τ
i
2
]
,
and
wherein the fractional number of intervals include partial intervals that remain inside the first or second time windows.
3. The method of claim 2 , wherein the method further comprises:
providing, by the at least one device, an indication of the cardiac activity feature representations of the mammal over the period of time; and
wherein the physiological event is selected from the onset of acquired epilepsy, a seizure, and sudden unexpected death in epilepsy (SUDEP).
4. The method of claim 2 , wherein τ 1 is greater than or equal to 100 ms.
5. The method of claim 2 , wherein τ 1 is greater than or equal to 1 second.
6. The method of claim 2 , wherein τ 1 is about 0.5 seconds and τ 2 is about 1 second.
7. The method of claim 2 , wherein τ 1 is from about 5 to about 20 seconds and τ 2 is about 15 to about 30 seconds.
8. The method of claim 2 , wherein the physiological data is ECG data and wherein the reference points correspond to one or more morphological feature representations of an ECG of the mammal.
9. The method of claim 8 , wherein the morphological feature representations are selected from a T-wave peak, a P-wave peak, a R-wave peak, an onset of a P-wave, an onset of Q, and an end of a T-wave.
10. The method of claim 2 , wherein at least one of the two or more reference points are representative of at least one feature of a QRS complex or at least one feature within a QRS complex of the mammal.
11. The method of claim 2 , wherein the reference points are R wave peaks.
12. The method of claim 2 , wherein the one or more intervals include at least one RR interval.
13. The method of claim 2 , wherein each of the one or more cardiac activity feature representations describe a single heartbeat of the mammal.
14. The method of claim 1 , wherein the brain activity data is EEG data and the EEG data is EEG line length data.
15. The method of claim 1 , wherein the mammal is human, mouse, or rat.
16. The method of claim 1 , wherein the periphery data is selected from the group consisting of cardiac data, muscle movement data, nerve activity data, pupil dilation data, and body movement data.
17. The method of claim 16 , wherein the cardiac data is selected from a group consisting of electrocardiogram (ECG) data, blood pressure data, and photoplethysmography (PPG) data.
18. The method of claim 16 , wherein the muscle movement data is selected from electromyography (EMG) data, and limb or hand acceleration data.