Removing eletrophysicologic artifacts from a magnetic resonance imaging system

A system for monitoring a subject's brain, the system comprising:

• an input configured to receive electroencephalogram (“EEG”) data acquired from a subject;
• at least one processor configured to:
• i) receive the EEG data from the input;
• ii) apply a regression model that defines a ballistocardiogram signal using a harmonic representation to the EEG data to define a cost function;
• iii) perform an iterative optimization process to estimate regression parameters that reduce the cost function;
• v) determine, using the regression parameters, the ballistocardiogram signal;
• vi) generate a corrected EEG data by reducing the ballistocardiogram signal within the EEG data;
• an output configured to deliver a report representative of the corrected EEG data;
• and
• wherein the cost function is defined as:
• C
• ⁡
• (
• ω
• ,
• β
• ,
• α
• |
• y
• )
• =
• T
• ⁢
• ⁢
• log
• ⁡
• (
• S
• T
• T
• )
• +
• log
• ⁡
• (
• det
• ⁡
• (
• Q
• )
• )
• +
• N
• ;
• wherein ST=(y−Z(ω)β)TQ−1(y−Z(ω)β) for an observed T×1 data vector of the time-series EEG data, y, N is a number of harmonics in a spectrum of the time-series EEG data, and ω (rad/sec) is a fundamental frequency defining a harmonic template, α and β are the estimated parameters, Q is a covariance matrix and Z is a harmonic model matrix.

A system for monitoring a subject's brain, the system comprising:

• an input configured to receive electroencephalogram (“EEG”) data acquired from a subject;
• at least one processor configured to:
• i) receive the EEG data from the input;
• ii) apply a regression model that defines a ballistocardiogram signal using a harmonic representation to the EEG data to define a cost function;
• iii) perform an iterative optimization process to estimate regression parameters that reduce the cost function;
• v) determine, using the regression parameters, the ballistocardiogram signal;
• vi) generate a corrected EEG data by reducing the ballistocardiogram signal within the EEG data;
• an output configured to deliver a report representative of the corrected EEG data;
• and
• wherein the cost function is defined as:
• C
• ⁡
• (
• ω
• ,
• β
• ,
• α
• |
• y
• )
• =
• T
• ⁢
• ⁢
• log
• ⁡
• (
• S
• T
• T
• )
• +
• log
• ⁡
• (
• det
• ⁡
• (
• Q
• )
• )
• +
• N
• ;
• where is ST=(y−Z(ω)β)TQ−1(y−Z(ω)β) for an observed T×1 data vector of the EEG data, y, N is a number of harmonics in a spectrum of the EEG data as a time-series EEG signal, and ω (rad/sec) is a fundamental frequency defining a harmonic template, α and β are the estimated parameters, Q is a covariance matrix and Z is a harmonic model matrix.

A method for reducing artifacts in an electrophysiologic signal, the method comprising:

• receiving a time-series electrophysiologic signal acquired from a subject;
• providing a regression model that defines an interference signal caused by periodic artifacts using a harmonic representation;
• applying the regression model using the time-series electrophysiologic signal to define a cost function;
• performing an iterative optimization process to estimate regression parameters that minimize the cost function;
• determining, using the regression parameters, the interference signal;
• generating a corrected time-series electrophysiologic signal by reducing the interference signal;
• wherein the cost function is defined as:
• C
• ⁡
• (
• ω
• ,
• β
• ,
• α
• |
• y
• )
• =
• T
• ⁢
• ⁢
• log
• ⁡
• (
• S
• T
• T
• )
• +
• log
• ⁡
• (
• det
• ⁡
• (
• Q
• )
• )
• +
• N
• ;
• where is ST=(y−Z(ω)β)TQ−1(y−Z(ω)β) for an observed T×1 data vector of the electrophysiologic signal, y, N is a number of harmonics in a spectrum of the time-series electrophysiologic signal, and ω (rad/sec) is a fundamental frequency defining a harmonic template, α and β are the estimated parameters, Q is a covariance matrix and Z is a harmonic model matrix;
• and
• wherein the electrophysiologic signal is an electroencephalogram (“EEG”) signal.