Analysis of brain patterns using temporal measures
A set of brain data representing a time series of neurophysiologic activity acquired by spatially distributed sensors arranged to detect neural signaling of a brain (such as by the use of magnetoencephalography) is obtained. The set of brain data is processed to obtain a dynamic brain model based on a set of statistically-independent temporal measures, such as partial cross correlations, among groupings of different time series within the set of brain data. The dynamic brain model represents interactions between neural populations of the brain occurring close in time, such as with zero lag, for example. The dynamic brain model can be analyzed to obtain the neurophysiologic assessment of the brain. Data processing techniques may be used to assess structural or neurochemical brain pathologies.
1. A system for classifying neurophysiologic activity of a first subject, the system comprising:
a data input configured to receive brain activity data corresponding to an idle state of the brain of the first subject, the brain activity data representing a time series of neurophysiologic activity acquired by a sensor system arranged to detect spatial and temporal neural signaling in the subject utilizing magnetic fields produced in a multiplicity of brain regions; and
a processor communicatively coupled to the data input, and programmed to:
process each set of brain activity data to produce a corresponding dynamic model of neural activity representing time-dependent coupling between neural populations of the brain of the first subject, including:
processing the brain activity data to produce a prewhitened time series;
computing partial cross correlations of the prewhitened time series to produce estimates of strength and sign of signaling between groups of the multiplicity of brain regions representing interactions of neural populations, based on an analysis of covariance of at least one type of partial cross correlations selected from the group consisting of: (a) positive partial cross correlations of the partial cross correlations of the prewhitened time series, and (b) negative partial cross correlations of the partial cross correlations of the prewhitened time series;
performing a classification of the partial cross correlations to produce a measure of correlation of the brain activity data to validated reference data corresponding to a plurality of different neurophysiologic conditions.
2. The system of claim 1 , wherein the plurality of different neurophysiologic conditions includes at least two conditions selected from the group consisting of: a normal condition, Alzheimer's Disease, pre-dementia syndrome, mild cognitive impairment, schizophrenia, Sjögren's Syndrome, alcoholism, alcohol impairment, fetal alcohol syndrome, multiple sclerosis, Parkinson's Disease, bipolar disorder, traumatic brain injury, depression, an autoimmune disorder, a neurodegenerative disorder, pain, a disease affecting the central nervous system, or any combination thereof.
3. The system of claim 1 , wherein the processor is programmed to apply an autoregressive integrative moving average-based algorithm to produce the prewhitened time series.
4. The system of claim 1 , wherein the partial cross correlations are computed to produce estimates of strength and sign of direct short-term signaling between the pairs of the multiplicity of sensors occurring within time windows of about 50 milliseconds.
5. The system of claim 1 , wherein the partial cross correlations are computed to produce estimates of strength and sign of substantially simultaneous signaling between the pairs of the multiplicity of sensors within a time window of 1 millisecond.
6. The system of claim 1 , wherein the processor is further programmed to apply a thresholding function to the estimates of strength and sign of signaling between pairs of the multiplicity of sensors to produce the first dynamic model of neural activity.
7. The system of claim 1 , wherein the processor is programmed to determine a relevant subset of the partial cross correlations that consists of certain partial cross correlations relevant for performing the classification.
8. The system of claim 7 , wherein the processor is programmed to determine the relevant subset of the partial cross correlations utilizing a leave-one-out algorithm wherein 100% classification is required for any one of the partial cross correlations to be deemed relevant.
9. The system of claim 7 , wherein the processor is programmed to determine the relevant subset of the partial cross correlations utilizing a genetic algorithm.
10. The system of claim 1 , wherein the processor is programmed to develop a neurophysiologic template based on the classification.
11. The system of claim 1 , wherein the dynamic model of neural activity is a first dynamic brain model of the first subject based on the brain activity data, which is taken at a first time, and wherein system is further configured to produce a different dynamic model of neural activity based on a second dynamic brain model of the first subject obtained from a second set of brain activity data taken at a second time that is different from the first time; and
wherein the processor is programmed to analyze a potential change in neurophysiology of the first subject between the first time and the second time based on a comparison of the first dynamic brain model and the second dynamic brain model.
12. The system of claim 1 , wherein the groups of the multiplicity of brain regions between which the estimates of strength and sign of the signaling are computed are pairs of brain regions, such that the interactions of neural populations are pair-wise interactions of neural populations.
13. The system of claim 1 , wherein the processor is programmed to produce the dynamic model based on the brain activity data including data representing the time series of neurophysiologic activity-corresponding to less than or equal to one minute of eye fixation activity.
14. The system of claim 1 , wherein the sensor system arranged to detect spatial and temporal neural signaling in the subject utilizing magnetic fields produced in a multiplicity of brain regions is a magnetoencephalogram instrument having a plurality of sensors arranged around the brain of the first subject, with each sensor gathering brain activity data from a corresponding one of the multiplicity of brain regions.
15. A non-transitory computer-readable medium comprising instructions that are adapted to cause a computer system to:
receive sets of brain activity data corresponding to an idle state of the brain of a first subject, each set representing a time series of neurophysiologic activity acquired by a sensor system arranged to detect spatial and temporal neural signaling in the subject utilizing magnetic fields produced in a multiplicity of brain regions;
process each set of brain activity data to produce a corresponding dynamic model of neural activity representing time-dependent coupling between neural populations of the brain of the first subject, including:
processing the brain activity data to produce a prewhitened time series;
computing partial cross correlations of the prewhitened time series to produce estimates of strength and sign of signaling between groups of the multiplicity of brain regions representing interactions of neural populations, based on an analysis of covariance of at least one type of partial cross correlations selected from the group consisting of: (a) positive partial cross correlations of the partial cross correlations of the prewhitened time series, and (b) negative partial cross correlations of the partial cross correlations of the prewhitened time series; and
perform a classification of the partial cross correlations to produce a measure of correlation of the brain activity data to validated reference data corresponding to a plurality of different neurophysiologic conditions.