Classes
class	RtAveragingWorker
	Background worker thread that accumulates and averages epochs in real time. More...
class	RtAveraging
	Controller that manages RtAveragingWorker for online epoch averaging with baseline correction. More...
class	RtConnectivityWorker
	Background worker thread that computes functional connectivity metrics in real time. More...
class	RtConnectivity
	Controller that manages RtConnectivityWorker for online connectivity computation. More...
struct	RtCovComputeResult
	Bundled output of a real-time covariance computation step containing the covariance matrix and sample count. More...
class	RtCov
	Controller that manages background covariance matrix estimation from streaming data. More...
struct	FilterObject
	Lightweight filter configuration holding kernel coefficients and overlap-add state for one channel. More...
class	FilterOverlapAdd
	Applies FIR filtering via FFT-based overlap-add convolution for continuous data streams. More...
class	RtHpiWorker
	Background worker thread that runs continuous HPI coil localization. More...
class	RtHpi
	Controller that manages RtHpiWorker for continuous head position tracking. More...
struct	RtInvOpInput
	Input bundle for the real-time inverse operator worker containing noise covariance, forward solution, and settings. More...
class	RtInvOpWorker
	Background worker thread that recomputes the MNE inverse operator when covariance updates arrive. More...
class	RtInvOp
	Controller that manages RtInvOpWorker for online inverse operator updates. More...
class	RtNoiseWorker
	Background worker that computes a noise power spectral density estimate from accumulated data blocks. More...
class	RtNoise
	Controller that manages RtNoiseWorker for real-time noise spectrum estimation. More...

Functions
DSPSHARED_EXPORT QList< Eigen::MatrixXi >	toEventMatrix (QMap< int, QList< QPair< int, double > > > mapTriggers)
DSPSHARED_EXPORT QMap< int, QList< QPair< int, double > > >	detectTriggerFlanksMax (const Eigen::MatrixXd &data, const QList< int > &lTriggerChannels, int iOffsetIndex, double dThreshold, bool bRemoveOffset, int iBurstLengthSamp=100)
DSPSHARED_EXPORT QList< QPair< int, double > >	detectTriggerFlanksMax (const Eigen::MatrixXd &data, int iTriggerChannelIdx, int iOffsetIndex, double dThreshold, bool bRemoveOffset, int iBurstLengthSamp=100)
DSPSHARED_EXPORT QMap< int, QList< QPair< int, double > > >	detectTriggerFlanksGrad (const Eigen::MatrixXd &data, const QList< int > &lTriggerChannels, int iOffsetIndex, double dThreshold, bool bRemoveOffset, const QString &type, int iBurstLengthSamp=100)
DSPSHARED_EXPORT QList< QPair< int, double > >	detectTriggerFlanksGrad (const Eigen::MatrixXd &data, int iTriggerChannelIdx, int iOffsetIndex, double dThreshold, bool bRemoveOffset, const QString &type, int iBurstLengthSamp=100)
DSPSHARED_EXPORT bool	filterFile (QIODevice &pIODevice, QSharedPointer< FIFFLIB::FiffRawData > pFiffRawData, int type, double dCenterfreq, double dBandwidth, double dTransition, double dSFreq, int iOrder=4096, int designMethod=UTILSLIB::FilterKernel::m_designMethods.indexOf(UTILSLIB::FilterParameter("Cosine")), const Eigen::RowVectorXi &vecPicks=Eigen::RowVectorXi(), bool bUseThreads=true)
DSPSHARED_EXPORT bool	filterFile (QIODevice &pIODevice, QSharedPointer< FIFFLIB::FiffRawData > pFiffRawData, const UTILSLIB::FilterKernel &filterKernel, const Eigen::RowVectorXi &vecPicks=Eigen::RowVectorXi(), bool bUseThreads=false)
DSPSHARED_EXPORT Eigen::MatrixXd	filterData (const Eigen::MatrixXd &matData, int type, double dCenterfreq, double dBandwidth, double dTransition, double dSFreq, int iOrder=1024, int designMethod=UTILSLIB::FilterKernel::m_designMethods.indexOf(UTILSLIB::FilterParameter("Cosine")), const Eigen::RowVectorXi &vecPicks=Eigen::RowVectorXi(), bool bUseThreads=true, bool bKeepOverhead=false)
DSPSHARED_EXPORT Eigen::MatrixXd	filterData (const Eigen::MatrixXd &matData, const UTILSLIB::FilterKernel &filterKernel, const Eigen::RowVectorXi &vecPicks=Eigen::RowVectorXi(), bool bUseThreads=true, bool bKeepOverhead=false)
DSPSHARED_EXPORT Eigen::MatrixXd	filterDataBlock (const Eigen::MatrixXd &matData, const Eigen::RowVectorXi &vecPicks, const UTILSLIB::FilterKernel &filterKernel, bool bUseThreads=true)
DSPSHARED_EXPORT void	filterChannel (FilterObject &channelDataTime)
DSPSHARED_EXPORT FIFFLIB::FiffEvoked	computeFilteredAverage (const FIFFLIB::FiffRawData &raw, const Eigen::MatrixXi &matEvents, float fTMinS, float fTMaxS, qint32 eventType, bool bApplyBaseline, float fTBaselineFromS, float fTBaselineToS, const QMap< QString, double > &mapReject, const UTILSLIB::FilterKernel &filterKernel, const QStringList &lExcludeChs=QStringList(), const Eigen::RowVectorXi &vecPicks=Eigen::RowVectorXi())

Function Documentation

◆ computeFilteredAverage()

DSPSHARED_EXPORT FIFFLIB::FiffEvoked RTPROCESSINGLIB::computeFilteredAverage	(	const FIFFLIB::FiffRawData &	raw,
		const Eigen::MatrixXi &	matEvents,
		float	fTMinS,
		float	fTMaxS,
		qint32	eventType,
		bool	bApplyBaseline,
		float	fTBaselineFromS,
		float	fTBaselineToS,
		const QMap< QString, double > &	mapReject,
		const UTILSLIB::FilterKernel &	filterKernel,
		const QStringList &	lExcludeChs = QStringList(),
		const Eigen::RowVectorXi &	vecPicks = Eigen::RowVectorXi() )

Computes the filtered average for given fiff raw data. Reads epochs, filters each one, applies baseline correction and artifact rejection, then returns the averaged evoked response.

Parameters

[in]	raw	The raw data.
[in]	matEvents	The events provided in samples and event kinds.
[in]	fTMinS	The start time relative to the event in seconds.
[in]	fTMaxS	The end time relative to the event in seconds.
[in]	eventType	The event type.
[in]	bApplyBaseline	Whether to use baseline correction (mode=mean).
[in]	fTBaselineFromS	The start baseline correction time relative to the event in seconds.
[in]	fTBaselineToS	The end baseline correction time relative to the event in seconds.
[in]	mapReject	The thresholds per channel type to reject epochs.
[in]	filterKernel	The filter kernel to use when reading the fiff raw data.
[in]	lExcludeChs	List of channel names to exclude.
[in]	vecPicks	Which channels to pick.

Returns: The filtered, averaged evoked data.

◆ detectTriggerFlanksGrad() [1/2]

DSPSHARED_EXPORT QMap< int, QList< QPair< int, double > > > RTPROCESSINGLIB::detectTriggerFlanksGrad	(	const Eigen::MatrixXd &	data,
		const QList< int > &	lTriggerChannels,
		int	iOffsetIndex,
		double	dThreshold,
		bool	bRemoveOffset,
		const QString &	type,
		int	iBurstLengthSamp = 100 )

detectTriggerFlanksGrad detects flanks from a given data matrix in row wise order. This function uses a simple gradient to locate the triggers.

Parameters

[in]	data	the data used to find the trigger flanks.
[in]	lTriggerChannels	The indeces of the trigger channels.
[in]	iOffsetIndex	the offset index gets added to the found trigger flank index.
[in]	iThreshold	the gradient threshold value used to find the trigger flank.
[in]	bRemoveOffset	remove the first sample as offset.
[in]	type	detect rising or falling flank. Use "Rising" or "Falling" as input.
[in]	iBurstLengthMs	The length in samples which is skipped after a trigger was found.

Returns: This map holds the indices of the channels which are to be read from data. For each index/channel the found triggers and corresponding signal values are written to the value of the map.

◆ detectTriggerFlanksGrad() [2/2]

DSPSHARED_EXPORT QList< QPair< int, double > > RTPROCESSINGLIB::detectTriggerFlanksGrad	(	const Eigen::MatrixXd &	data,
		int	iTriggerChannelIdx,
		int	iOffsetIndex,
		double	dThreshold,
		bool	bRemoveOffset,
		const QString &	type,
		int	iBurstLengthSamp = 100 )

detectTriggerFlanksGrad detects flanks from a given data matrix in row wise order. This function uses a simple gradient to locate the triggers.

Parameters

[in]	data	the data used to find the trigger flanks.
[in]	iTriggerChannelIdx	the index of the trigger channel in the matrix.
[in]	iOffsetIndex	the offset index gets added to the found trigger flank index.
[in]	iThreshold	the gradient threshold value used to find the trigger flank.
[in]	bRemoveOffset	remove the first sample as offset.
[in]	type	detect rising or falling flank. Use "Rising" or "Falling" as input.
[in]	iBurstLengthMs	The length in samples which is skipped after a trigger was found.

Returns: This list holds the found trigger indices and corresponding signal values.

◆ detectTriggerFlanksMax() [1/2]

DSPSHARED_EXPORT QMap< int, QList< QPair< int, double > > > RTPROCESSINGLIB::detectTriggerFlanksMax	(	const Eigen::MatrixXd &	data,
		const QList< int > &	lTriggerChannels,
		int	iOffsetIndex,
		double	dThreshold,
		bool	bRemoveOffset,
		int	iBurstLengthSamp = 100 )

detectTriggerFlanks detects flanks from a given data matrix in row wise order. This function uses a simple maxCoeff function implemented by eigen to locate the triggers.

Parameters

[in]	data	the data used to find the trigger flanks.
[in]	lTriggerChannels	The indeces of the trigger channels.
[in]	iOffsetIndex	the offset index gets added to the found trigger flank index.
[in]	dThreshold	the signal threshold value used to find the trigger flank.
[in]	bRemoveOffset	remove the first sample as offset.
[in]	iBurstLengthMs	The length in samples which is skipped after a trigger was found.

Returns: This map holds the indices of the channels which are to be read from data. For each index/channel the found triggersand corresponding signal values are written to the value of the map.

◆ detectTriggerFlanksMax() [2/2]

DSPSHARED_EXPORT QList< QPair< int, double > > RTPROCESSINGLIB::detectTriggerFlanksMax	(	const Eigen::MatrixXd &	data,
		int	iTriggerChannelIdx,
		int	iOffsetIndex,
		double	dThreshold,
		bool	bRemoveOffset,
		int	iBurstLengthSamp = 100 )

detectTriggerFlanks detects flanks from a given data matrix in row wise order. This function uses a simple maxCoeff function implemented by eigen to locate the triggers.

Parameters

[in]	data	the data used to find the trigger flanks.
[in]	iTriggerChannelIdx	the index of the trigger channel in the matrix.
[in]	iOffsetIndex	the offset index gets added to the found trigger flank index.
[in]	dThreshold	the signal threshold value used to find the trigger flank.
[in]	bRemoveOffset	remove the first sample as offset.
[in]	iBurstLengthMs	The length in samples which is skipped after a trigger was found.

Returns: This list holds the found trigger indices and corresponding signal values.

◆ filterChannel()

void RTPROCESSINGLIB::filterChannel ( RTPROCESSINGLIB::FilterObject & channelDataTime )

This function is used to filter row-wise in parallel threads

Parameters

[in] channelDataTime The channel data to perform the filtering on.

Definition at line 378 of file rt_filter.cpp.

◆ filterData() [1/2]

DSPSHARED_EXPORT Eigen::MatrixXd RTPROCESSINGLIB::filterData	(	const Eigen::MatrixXd &	matData,
		const UTILSLIB::FilterKernel &	filterKernel,
		const Eigen::RowVectorXi &	vecPicks = Eigen::RowVectorXi(),
		bool	bUseThreads = true,
		bool	bKeepOverhead = false )

Calculates the filtered version of the raw input data based on a given list filters. The data needs to be present all at once. For continuous filtering via overlap add use the FilterOverlapAdd class.

Parameters

[in]	matData	The data which is to be filtered.
[in]	filterKernel	The list of filter kernels to use.
[in]	vecPicks	Channel indexes to filter. Default is filter all channels.
[in]	bUseThreads	Whether to use multiple threads. Default is set to true.
[in]	bKeepOverhead	Whether to keep the delayed part of the data after filtering. Default is set to false .

Returns: The filtered data in form of a matrix.

◆ filterData() [2/2]

DSPSHARED_EXPORT Eigen::MatrixXd RTPROCESSINGLIB::filterData	(	const Eigen::MatrixXd &	matData,
		int	type,
		double	dCenterfreq,
		double	dBandwidth,
		double	dTransition,
		double	dSFreq,
		int	iOrder = 1024,
		int	designMethod = UTILSLIB::FilterKernel::m_designMethods.indexOf(UTILSLIB::FilterParameter("Cosine")),
		const Eigen::RowVectorXi &	vecPicks = Eigen::RowVectorXi(),
		bool	bUseThreads = true,
		bool	bKeepOverhead = false )

Creates a user designed filter kernel and filters the raw input data. The data needs to be present all at once. For continuous filtering via overlap add use the FilterOverlapAdd class.

Parameters

[in]	matData	The data which is to be filtered.
[in]	type	The type of the filter: LPF, HPF, BPF, NOTCH (from enum FilterType).
[in]	dCenterfreq	The center of the frequency.
[in]	dBandwidth	The filter bandwidth. Ignored if FilterType is set to LPF,HPF. If NOTCH/BPF: bandwidth of stop-/passband.
[in]	dTransition	The transition band determines the width of the filter slopes (steepness).
[in]	dSFreq	The input data sampling frequency.
[in]	iOrder	Represents the order of the filter, the higher the higher is the stopband attenuation. Default is 1024 taps.
[in]	designMethod	The design method to use. Choose between Cosine and Tschebyscheff. Default is set to Cosine.
[in]	vecPicks	Channel indexes to filter. Default is filter all channels.
[in]	bUseThreads	Whether to use multiple threads. Default is set to true.
[in]	bKeepOverhead	Whether to keep the delayed part of the data after filtering. Default is set to false .

Returns: The filtered data in form of a matrix.

◆ filterDataBlock()

DSPSHARED_EXPORT Eigen::MatrixXd RTPROCESSINGLIB::filterDataBlock	(	const Eigen::MatrixXd &	matData,
		const Eigen::RowVectorXi &	vecPicks,
		const UTILSLIB::FilterKernel &	filterKernel,
		bool	bUseThreads = true )

Calculates the filtered version of the raw input data block. Always returns the data with half the filter length delay in the front and back.

Parameters

[in]	matData	The data which is to be filtered.
[in]	vecPicks	The used channel as index in RowVector.
[in]	filterKernel	The FilterKernel to filter the data with.
[in]	bUseThreads	Whether to use multiple threads.

Returns: The filtered data in form of a matrix with half the filter length delay in the front and back.

◆ filterFile() [1/2]

DSPSHARED_EXPORT bool RTPROCESSINGLIB::filterFile	(	QIODevice &	pIODevice,
		QSharedPointer< FIFFLIB::FiffRawData >	pFiffRawData,
		const UTILSLIB::FilterKernel &	filterKernel,
		const Eigen::RowVectorXi &	vecPicks = Eigen::RowVectorXi(),
		bool	bUseThreads = false )

Filters data from an input file based on an existing filter kernel and writes the filtered data to a pIODevice.

Parameters

[in]	pIODevice	The IO device to write to.
[in]	pFiffRawData	The fiff raw data object to read from.
[in]	filterKernel	The list of filter kernels to use.
[in]	vecPicks	Channel indexes to filter. Default is filter all channels.
[in]	bUseThreads	Whether to use multiple threads. Default is set to true.

Returns: Returns true if successful, false otherwise.

◆ filterFile() [2/2]

DSPSHARED_EXPORT bool RTPROCESSINGLIB::filterFile	(	QIODevice &	pIODevice,
		QSharedPointer< FIFFLIB::FiffRawData >	pFiffRawData,
		int	type,
		double	dCenterfreq,
		double	dBandwidth,
		double	dTransition,
		double	dSFreq,
		int	iOrder = 4096,
		int	designMethod = UTILSLIB::FilterKernel::m_designMethods.indexOf(UTILSLIB::FilterParameter("Cosine")),
		const Eigen::RowVectorXi &	vecPicks = Eigen::RowVectorXi(),
		bool	bUseThreads = true )

Creates a user designed filter kernel, filters data from an input file and writes the filtered data to a pIODevice.

Parameters

[in]	pIODevice	The IO device to write to.
[in]	pFiffRawData	The fiff raw data object to read from.
[in]	type	The type of the filter: LPF, HPF, BPF, NOTCH (from enum FilterType).
[in]	dCenterfreq	The center of the frequency.
[in]	dBandwidth	The filter bandwidth. Ignored if FilterType is set to LPF,HPF. If NOTCH/BPF: bandwidth of stop-/passband.
[in]	dTransition	The transition band determines the width of the filter slopes (steepness).
[in]	dSFreq	The input data sampling frequency.
[in]	iOrder	Represents the order of the filter, the higher the higher is the stopband attenuation. Default is 4096 taps.
[in]	designMethod	The design method to use. Choose between Cosine and Tschebyscheff. Default is set to Cosine.
[in]	vecPicks	Channel indexes to filter. Default is filter all channels.
[in]	bUseThreads	Whether to use multiple threads. Default is set to true.

Returns: Returns true if successful, false otherwise.

◆ toEventMatrix()

QList< MatrixXi > RTPROCESSINGLIB::toEventMatrix ( QMap< int, QList< QPair< int, double > > > mapTriggers )

Transforms QMap with stored information about events per stim channel to a list of event matrices.

Parameters

[in]	mapTriggers	The QMap to be transformed.
[in]	A	list of transformed Eigen matrices.

Definition at line 58 of file rt_detect_trigger.cpp.

Classes

Functions

Function Documentation

◆ computeFilteredAverage()

◆ detectTriggerFlanksGrad() [1/2]

◆ detectTriggerFlanksGrad() [2/2]

◆ detectTriggerFlanksMax() [1/2]

◆ detectTriggerFlanksMax() [2/2]

◆ filterChannel()

◆ filterData() [1/2]

◆ filterData() [2/2]

◆ filterDataBlock()

◆ filterFile() [1/2]

◆ filterFile() [2/2]

◆ toEventMatrix()