Systemic Influence on Hemodynamic Responses EHK-NIRS
Full project title: Systemic influence on characteristic hemodynamic responses measured with NIRS
Short project title: EHK-NIRS
Summary
Near infrared spectroscopy (NIRS) is a recently developed technique that can reveal hemodynamic and metabolic changes during cortical activation. Unlike established techniques, such as electroencephalography (EEG), functional magnetic resonance imaging (fMRI), and magnetoencephalography (MEG), NIRS could be more practical and user-friendly. NIRS has been used to study hemodynamic responses (changes of oxy- and deoxyhemoglobin) to cognitive, visual, and motor tasks. A big challenge when using NIRS is the classification of single trial data. Single trial classification requires improving the signal to noise ratio (SNR) and reducing false classifications that primarily stem from misclassification of physiological noise. To determine whether the recorded signal reflects local cortical activation or a global response of the cardiovascular system, it is essential to identify systemic influences.
The goal of this project is to investigate the influence of systemic parameters on hemodynamic responses measured with NIRS. Another aim is to design appropriate signal processing approaches to reduce the systemic influences in the recorded NIRS signals.
Publications
Cardiovascular responses after brisk finger movement and their dependency on the "eigenfrequency" of the baroreflex loop
The baroreflex is mainly involved in short-term blood pressure regulation and strongly influenced by activations of medullary circulation centres in the brain stem and higher brain centres. One important feature of the baroreflex is its strong preference for oscillations around 0.1 Hz, which can be seen as resonance or "eigenfrequency" (EF) of the control loop (so-called Mayer waves). In the present study we investigated beat-to-beat heart rate intervals (RRI) and arterial blood pressure (BP) changes after brisk finger movement and their relationship to the "eigenfrequency" determined by cross spectral analysis between RRI and arterial blood pressure time series of 17 healthy subjects. The analyses revealed significant correlations between BP response magnitude (r = 0.63, p < 0.01) respectively RRI response magnitude (r = 0.59, p < 0.05) and EF. This can be interpreted in such a way that subjects with a "high" EF (>0.10 Hz) elicit larger BP responses as well as larger RRI responses when compared to subjects with a "low" EF (<0.10 Hz).
For more details, see http://dx.doi.org/10.1016/j.neulet.2010.12.020. About the stability of phase-shifts between slow oscillations around 0.1 Hz in cardiovascular and cerebral systems
One important feature of the baroreflex loop is its strong preference for oscillations around 0.1 Hz. In this study, we investigated heart rate intervals, arterial blood pressure (BP), and prefrontal oxyhemoglobin changes during 5 min rest and during brisk finger movements in 19 healthy subjects. We analyzed the phase coupling around 0.1 Hz between cardiovascular and (de)oxyhemoglobin oscillations, using the cross-spectral method. The analyses revealed phase shifts for slow oscillations in BP and heart rate intervals between -10° and-118° (BP always leading). These phase shifts increased significantly (p < 0.01) in the movement session. The coupling between cardiovascular and oxyhemoglobin oscillations was less clear. Only 12 subjects demonstrated a phase coupling (COH^2 > 0.5) between oxyhemoglobin and BP oscillations. This may be explained by an overwhelming proportion of nonlinearity in cardiovascular and hemodynamic systems. The phase shifts between slow cardiovascular and hemodynamic oscillations are relatively stable subject-specific biometric features and could be of interest for person identification in addition to other biometric data. Slow BP-coupled oscillations in prefrontal oxyhemoglobin changes can seriously impair the detection of mentally induced hemodynamic changes in an optical brain–computer interface, a novel nonmuscular communication system.
For more details, see http://dx.doi.org/10.1109/TBME.2011.2134851.
Does conscious intention to perform a motor act depend on slow prefrontal(de)oxyhemoglobin oscillations in the resting brain?
Characteristically within the resting brain there are slow fluctuations (around 0.1 Hz) of EEG and NIRS-(de)oxyhemoglobin ([deoxy-Hb], [oxy-Hb]) signals. An interesting question is whether such slow oscillations can be related to the intention to perform a motor act. To obtain an answer we analyzed continuous blood pressure (BP), heart rate (HR), prefrontal [oxy-Hb], [deoxy-Hb] and EEG signals over sensorimotor areas in 10 healthy subjects during 5 min of rest and during 10 min of voluntary finger movements. Analyses of prefrontal [oxy-Hb]/[deoxy-Hb] oscillations around 0.1 Hz and central EEG band power changes in the beta (alpha) band revealed that the positive [oxy-Hb] peaks preceded the central EEG beta (alpha) power peak by 3.6 +/- 0.9 s in the majority of subjects. A similar relationship between prefrontal [oxy-Hb] and central EEG beta power was found during voluntary movements whereby the post movement beta power increase (beta rebound) is known to coexist with a decreased excitability of cortico-spinal neurons. Therefore, we speculate that the beta power increase about 3 s after slow fluctuating [oxy-Hb] peaks during rest is indicative for a slow excitability change of central motor cortex neurons. This work provides the first evidence that initiation of finger movements at free will in relatively constant intervals around 10 s could be temporally related to slow oscillations of prefrontal [oxy-Hb] and autonomic blood pressure in the resting brain.
For more details, see http://dx.doi.org/10.1016/j.neulet.2011.12.025.
Single-trial classification of antagonistic oxyhemoglobin responses during mental arithmetic
Near-infrared spectroscopy (NIRS) is a noninvasive optical technique that can be used for brain-computer interfaces (BCIs) systems. A common challenge for BCIs is a stable and reliable classification of single-trial data, especially for cognitive (mental) tasks. With antagonistic activation pattern, recently found for mental arithmetic (MA) tasks, an improved online classification for optical BCIs using MA should become possible. For this investigation, we used the data of a previous study where we found antagonistic activation patterns (focal bilateral increase of [oxy-Hb] in the dorsolateral prefrontal cortex in parallel with a [oxy-Hb] decrease in the medial area of the anterior prefrontal cortex) in eight subjects. We used the [oxy-Hb] responses to search for the best antagonistic feature combination and compared it to individual features from the same regions. In addition, we investigated the use of antagonistic [deoxy-Hb], total hemoglobin [Hbtot] and pairs of [oxy-Hb] and [deoxy-Hb] features as well as the existence of a group-related feature set. Our results indicate that the use of the antagonistic [oxy-Hb] features significantly increases the classification accuracy from 63.3 to 79.7%. These results support the hypothesis that antagonistic hemodynamic response patterns are a suitable control strategy for optical BCI, and that only two prefrontal NIRS channels are needed for good performance.
For more details, see http://dx.doi.org/10.1007/s11517-011-0792-5.
Focal frontal (de)oxyhemoglobin responses during simple arithmetic
Near-infrared spectroscopy (NIRS) is a functional brain imaging method able to study hemodynamic changes during cortical activation. We studied the changes of oxy- and deoxyhemoglobin ([oxy-Hb], [deoxy-Hb]) with a 52-channel NIRS system during simple mental arithmetic in ten healthy volunteers over the prefrontal cortex. We found that eight of the ten subjects showed a relative focal bilateral increase of [oxy-Hb] in the dorsolateral prefrontal cortex (DLPFC) in parallel with a decrease in the medial area of the anterior prefrontal cortex (APFC). The [oxy-Hb] response in left DLPFC and APFC was significant, while the [deoxy-Hb] response was clearly smaller and not significant. These observations were discussed within the context of "focal activation/surround deactivation".
For more details, see http://dx.doi.org/10.1016/j.ijpsycho.2010.03.013.
Investigation of hemodynamic responses during simple arithmetic -- annotations for the use as control signal for optical brain-computer interface (oBCI) applications
During the past decades, several groups have dedicated research on the field of brain-computer interaction as a communication and control support for paralyzed patients. Brain-computer interface (BCI) systems are normally based on the EEG. Beside EEG, such systems have also been realized with MEG and recently with fMRI methods. However, the use of these techniques is restricted due to system size and ambient requirements. As a more promising method the near-infrared spectroscopy (NIRS) technique can also be used for future BCIs based on hemodynamic responses. Besides motor tasks mainly cognitive tasks, in particular mental arithmetic (MA) operations in form of one-digit tasks are used. Several NIRS studies have demonstrated the implication of the prefrontal cortex (PFC) during MA, and found characteristic hemodynamic responses, but most of them used only one or two NIRS channels. To obtain better insight in PFC activation during MA performance and the use of the response for BCI applications, we studied hemodynamic responses with a multichannel NIRS system during simple MA, which is to our knowledge the first study using 52 channels.
For more details, see TUGOnline.
Neural correlates of the execution and inhibition of well learned foot and finger movements: an NIRS study
Adequate behaviour requires the control (execution/inhibition) of learned programs in a situational context. Inspired by the fMRI-work of Hummel the aim of this study was to investigate the appropriate control of acquired and memorized motor programs (to our knowledge the first time) with the fNIRS method.
For more details, see TUGOnline.
Cortical effects of BCI training measured with fNIRS
This study investigates the cortical training effects using a 2-class motor imagery (MI) based BCI. Twelve subjects were trained to use right hand or feet MI to control a cursor on a screen. The feedback was calculated by using features based on the EEG. To assess which areas are involved in the training, and how activity in these areas changes over time, three fNIRS measurements were applied before, in between and after the training. The statistical analysis of the measurement revealed that significant activity changes in the involved areas during the training can be found and that they occur accordingly to the task.
For more details, see TUGOnline.
The Hybrid BCI
Self-activation is an important factor for brain-computer interface (BCI) systems to become more practical and user-friendly devices. This means that the user should be able to switch on or off the system autonomously. In this work, we investigate the realization of an asynchronous "hybrid" BCI by combining near infrared spectroscopy (NIRS) with steady-state visual evoked potentials (SSVEP). Therefore, we used a one channel NIRS system developed by our group to turn on and off an electrical hand orthosis controlled by SSVEP.
For more details, see publication at Frontiers . On the removal of physiological artifacts from fNIRS
In the present study we report on the reduction of physiological rhythms in hemodynamic signals recorded with functional near-infrared spectroscopy (fNIRS). We investigated the use of two different signal processing approaches (independent component analysis [ICA] and transfer function models [TFs]) to reduce the influence of respiratory and blood pressure rhythms (Mayer waves) on the hemodynamic responses. The results show that both approaches reduce the artifactual influences equivalently well. However, TFs do so with significantly lower impact (p < 0.01) on none artifactual signal components then ICA.For more details, see TUGOnline.
The baroreflex is mainly involved in short-term blood pressure regulation and strongly influenced by activations of medullary circulation centres in the brain stem and higher brain centres. One important feature of the baroreflex is its strong preference for oscillations around 0.1 Hz, which can be seen as resonance or "eigenfrequency" (EF) of the control loop (so-called Mayer waves). In the present study we investigated beat-to-beat heart rate intervals (RRI) and arterial blood pressure (BP) changes after brisk finger movement and their relationship to the "eigenfrequency" determined by cross spectral analysis between RRI and arterial blood pressure time series of 17 healthy subjects. The analyses revealed significant correlations between BP response magnitude (r = 0.63, p < 0.01) respectively RRI response magnitude (r = 0.59, p < 0.05) and EF. This can be interpreted in such a way that subjects with a "high" EF (>0.10 Hz) elicit larger BP responses as well as larger RRI responses when compared to subjects with a "low" EF (<0.10 Hz).
For more details, see http://dx.doi.org/10.1016/j.neulet.2010.12.020. About the stability of phase-shifts between slow oscillations around 0.1 Hz in cardiovascular and cerebral systems
One important feature of the baroreflex loop is its strong preference for oscillations around 0.1 Hz. In this study, we investigated heart rate intervals, arterial blood pressure (BP), and prefrontal oxyhemoglobin changes during 5 min rest and during brisk finger movements in 19 healthy subjects. We analyzed the phase coupling around 0.1 Hz between cardiovascular and (de)oxyhemoglobin oscillations, using the cross-spectral method. The analyses revealed phase shifts for slow oscillations in BP and heart rate intervals between -10° and-118° (BP always leading). These phase shifts increased significantly (p < 0.01) in the movement session. The coupling between cardiovascular and oxyhemoglobin oscillations was less clear. Only 12 subjects demonstrated a phase coupling (COH^2 > 0.5) between oxyhemoglobin and BP oscillations. This may be explained by an overwhelming proportion of nonlinearity in cardiovascular and hemodynamic systems. The phase shifts between slow cardiovascular and hemodynamic oscillations are relatively stable subject-specific biometric features and could be of interest for person identification in addition to other biometric data. Slow BP-coupled oscillations in prefrontal oxyhemoglobin changes can seriously impair the detection of mentally induced hemodynamic changes in an optical brain–computer interface, a novel nonmuscular communication system.
For more details, see http://dx.doi.org/10.1109/TBME.2011.2134851.
Does conscious intention to perform a motor act depend on slow prefrontal(de)oxyhemoglobin oscillations in the resting brain?
Characteristically within the resting brain there are slow fluctuations (around 0.1 Hz) of EEG and NIRS-(de)oxyhemoglobin ([deoxy-Hb], [oxy-Hb]) signals. An interesting question is whether such slow oscillations can be related to the intention to perform a motor act. To obtain an answer we analyzed continuous blood pressure (BP), heart rate (HR), prefrontal [oxy-Hb], [deoxy-Hb] and EEG signals over sensorimotor areas in 10 healthy subjects during 5 min of rest and during 10 min of voluntary finger movements. Analyses of prefrontal [oxy-Hb]/[deoxy-Hb] oscillations around 0.1 Hz and central EEG band power changes in the beta (alpha) band revealed that the positive [oxy-Hb] peaks preceded the central EEG beta (alpha) power peak by 3.6 +/- 0.9 s in the majority of subjects. A similar relationship between prefrontal [oxy-Hb] and central EEG beta power was found during voluntary movements whereby the post movement beta power increase (beta rebound) is known to coexist with a decreased excitability of cortico-spinal neurons. Therefore, we speculate that the beta power increase about 3 s after slow fluctuating [oxy-Hb] peaks during rest is indicative for a slow excitability change of central motor cortex neurons. This work provides the first evidence that initiation of finger movements at free will in relatively constant intervals around 10 s could be temporally related to slow oscillations of prefrontal [oxy-Hb] and autonomic blood pressure in the resting brain.
For more details, see http://dx.doi.org/10.1016/j.neulet.2011.12.025.
Single-trial classification of antagonistic oxyhemoglobin responses during mental arithmetic
Near-infrared spectroscopy (NIRS) is a noninvasive optical technique that can be used for brain-computer interfaces (BCIs) systems. A common challenge for BCIs is a stable and reliable classification of single-trial data, especially for cognitive (mental) tasks. With antagonistic activation pattern, recently found for mental arithmetic (MA) tasks, an improved online classification for optical BCIs using MA should become possible. For this investigation, we used the data of a previous study where we found antagonistic activation patterns (focal bilateral increase of [oxy-Hb] in the dorsolateral prefrontal cortex in parallel with a [oxy-Hb] decrease in the medial area of the anterior prefrontal cortex) in eight subjects. We used the [oxy-Hb] responses to search for the best antagonistic feature combination and compared it to individual features from the same regions. In addition, we investigated the use of antagonistic [deoxy-Hb], total hemoglobin [Hbtot] and pairs of [oxy-Hb] and [deoxy-Hb] features as well as the existence of a group-related feature set. Our results indicate that the use of the antagonistic [oxy-Hb] features significantly increases the classification accuracy from 63.3 to 79.7%. These results support the hypothesis that antagonistic hemodynamic response patterns are a suitable control strategy for optical BCI, and that only two prefrontal NIRS channels are needed for good performance.
For more details, see http://dx.doi.org/10.1007/s11517-011-0792-5.
Focal frontal (de)oxyhemoglobin responses during simple arithmetic
Near-infrared spectroscopy (NIRS) is a functional brain imaging method able to study hemodynamic changes during cortical activation. We studied the changes of oxy- and deoxyhemoglobin ([oxy-Hb], [deoxy-Hb]) with a 52-channel NIRS system during simple mental arithmetic in ten healthy volunteers over the prefrontal cortex. We found that eight of the ten subjects showed a relative focal bilateral increase of [oxy-Hb] in the dorsolateral prefrontal cortex (DLPFC) in parallel with a decrease in the medial area of the anterior prefrontal cortex (APFC). The [oxy-Hb] response in left DLPFC and APFC was significant, while the [deoxy-Hb] response was clearly smaller and not significant. These observations were discussed within the context of "focal activation/surround deactivation".
For more details, see http://dx.doi.org/10.1016/j.ijpsycho.2010.03.013.
Investigation of hemodynamic responses during simple arithmetic -- annotations for the use as control signal for optical brain-computer interface (oBCI) applications
During the past decades, several groups have dedicated research on the field of brain-computer interaction as a communication and control support for paralyzed patients. Brain-computer interface (BCI) systems are normally based on the EEG. Beside EEG, such systems have also been realized with MEG and recently with fMRI methods. However, the use of these techniques is restricted due to system size and ambient requirements. As a more promising method the near-infrared spectroscopy (NIRS) technique can also be used for future BCIs based on hemodynamic responses. Besides motor tasks mainly cognitive tasks, in particular mental arithmetic (MA) operations in form of one-digit tasks are used. Several NIRS studies have demonstrated the implication of the prefrontal cortex (PFC) during MA, and found characteristic hemodynamic responses, but most of them used only one or two NIRS channels. To obtain better insight in PFC activation during MA performance and the use of the response for BCI applications, we studied hemodynamic responses with a multichannel NIRS system during simple MA, which is to our knowledge the first study using 52 channels.
For more details, see TUGOnline.
Neural correlates of the execution and inhibition of well learned foot and finger movements: an NIRS study
Adequate behaviour requires the control (execution/inhibition) of learned programs in a situational context. Inspired by the fMRI-work of Hummel the aim of this study was to investigate the appropriate control of acquired and memorized motor programs (to our knowledge the first time) with the fNIRS method.
For more details, see TUGOnline.
Cortical effects of BCI training measured with fNIRS
This study investigates the cortical training effects using a 2-class motor imagery (MI) based BCI. Twelve subjects were trained to use right hand or feet MI to control a cursor on a screen. The feedback was calculated by using features based on the EEG. To assess which areas are involved in the training, and how activity in these areas changes over time, three fNIRS measurements were applied before, in between and after the training. The statistical analysis of the measurement revealed that significant activity changes in the involved areas during the training can be found and that they occur accordingly to the task.
For more details, see TUGOnline.
The Hybrid BCI
Self-activation is an important factor for brain-computer interface (BCI) systems to become more practical and user-friendly devices. This means that the user should be able to switch on or off the system autonomously. In this work, we investigate the realization of an asynchronous "hybrid" BCI by combining near infrared spectroscopy (NIRS) with steady-state visual evoked potentials (SSVEP). Therefore, we used a one channel NIRS system developed by our group to turn on and off an electrical hand orthosis controlled by SSVEP.
For more details, see publication at Frontiers . On the removal of physiological artifacts from fNIRS
In the present study we report on the reduction of physiological rhythms in hemodynamic signals recorded with functional near-infrared spectroscopy (fNIRS). We investigated the use of two different signal processing approaches (independent component analysis [ICA] and transfer function models [TFs]) to reduce the influence of respiratory and blood pressure rhythms (Mayer waves) on the hemodynamic responses. The results show that both approaches reduce the artifactual influences equivalently well. However, TFs do so with significantly lower impact (p < 0.01) on none artifactual signal components then ICA.For more details, see TUGOnline.
