Keywords

1 Introduction

There are various correlations between electrophysiological and hemodynamic responses in stroke, epilepsy, brain tumor, dementia, rehabilitation process, etc.

There are two main clinical approaches for cerebral vascular disease, especially stroke. The first is measurement of cerebrovascular reserve capacity, evaluation of which is very important to treat cerebral infarction. The other approach involves determination of brain plasticity, which is an indicator of rehabilitation after stroke [1].

Cerebral blood flow and metabolism can be measured by nuclear medicine studies, such as single-photon emission computed tomography (SPECT) and positron emission tomography (PET), as well as by perfusion CT, MRI, and near-infrared spectroscopy (NIRS). These methods have been established for diagnosing and assessing cerebrovascular disease. Some reports have indicated that brain plasticity can be evaluated by diffusion-weighted MRI or fMRI [2].

However, hemodynamics in cerebrovascular disease as assessed by these imaging studies only indirectly reflects brain function.

On the other hand, electrophysiological studies like electroencephalography (EEG) more directly reflect neural activity. It is well known that the slow wave activity in EEG occurs in ischemic areas of the brain. On the other hand, scalp-recorded EEGs are greatly affected by the skull itself. Electrophysiological signal in small infarct lesions may be buried within normal signal findings in surrounding areas and go undetected. Thus, routine EEGs are problematic because of low spatial resolution and a lack of objective indices for quantitative measurement of brain function. Magnetoencephalography (MEG) quantitatively measures magnetic activity in the brain and may overcome some of these limitations of EEG. This chapter mainly describes the relations between magnetic responses and cerebral ischemic changes.

MEG is now performed with devices that can cover the entire head, thus providing improved spatial resolution. MEG was first used in clinical research to assess brain function and search for epileptic foci preoperatively in patients undergoing neurosurgery. The usefulness of MEG for brain function evaluation in other diseases has also been frequently reported, but its clinical use in diseases other than epilepsy has been sparse, so it is still unclear what level of scientific evidence exists for its utility outside of epilepsy.

Therefore, trends in clinical research of MEG in ischemic cerebral disease were examined in a literature search to evaluate the level of scientific evidence for using MEG in clinical evaluation.

2 Methods

A PubMed database literature search (http://www.ncbi.nlm.nih.gov/pubmed/) was conducted with the keywords (stroke or cerebral ischemia) and (MEG or magnetoencephalography) for publications dated between January 1990 and July 2014. A total of 62 papers were retrieved from this search regarding MEG based on the titles, 28 of these papers were reviewed based on their abstract contents, and then, focusing on the level of evidence, 14 papers that measured cerebral blood flow (CBF) or used controlled comparisons were selected and reviewed.

Although it was difficult to directly compare these reports due to the nonuniformity of presentation and varied ischemic conditions, some common results are described.

3 Results 1

A total of 62 papers about MEG and cerebral ischemia were retrieved from the literature search. These were then limited to original manuscripts, and 28 papers were selected based on their abstract contents. These included reports about diagnostic criteria for cerebral ischemia and functional recovery/neuroplasticity after stroke. They were further divided into reports about spontaneous cerebral magnetic fields and evoked cerebral magnetic fields.

Evaluation of the use of MEG in functional recovery after stroke is reported separately, and some reports about the evaluation of ischemia are described.

3.1 Diagnostic Criteria for Cerebral Ischemia in MEG

  1. 1.

    Spontaneous magnetic fields: slow waves occur after stroke

    • Affected side

      • In areas surrounding lesions after stroke, MEG shows a decrease of high-frequency components (gamma band) and an increase of low-frequency delta (theta band) [3, 4].

      • Increased theta waves are seen in the temporoparietal area of the affected side [5] and correlate with misery perfusion [6].

      • Slow waves are seen on the affected side in transient ischemic attack (TIA) patients [7].

      • The power of delta waves is correlated with the National Institute of Health Stroke Scale (NIHSS) [8].

      • Slow wave components are correlated with decreased N-acetylaspartate (NAA) on MR spectroscopy [9].

      • Slow wave activity during re-buildup in areas of impaired circulation in moyamoya disease is seen in deep cortical sulci [10, 11].

    • Unaffected side

      • Increased slow waves are also seen on the unaffected side [4, 12].

      • Increased slow waves (delta) on the unaffected side may be clinically correlated with symptom improvement [2].

  2. 2.

    Evoked magnetic fields

    • SEF (somatosensory evoked fields)

      • In stroke cases, asymmetry of localization of equivalent current dipoles (ECDs) in N20m of the affected side [13], prolonged latency [14, 15], abnormal waveforms [16], and decreased [12, 17] or increased ECD strength [11, 15] are seen. This asymmetry is correlated with improvement of clinical symptoms [12, 18].

      • The SEF ECD component at N20m is correlated with the NIHSS [8].

      • In parallel with decreased CBF, N20m dipole moment decreases and P30m increases [19].

    • MEF (motor evoked fields)

      • Shifts in source position and latencies are seen [20].

      • Βeta band event-related desynchronization (ERD) is seen in motor areas and the ipsilateral hemisphere [12].

      • The amplitude of motor-related direct current (DC) signal (infra-slow MEG signal) decreases in the affected hemisphere [21].

    • AEF (auditory evoked fields)

      • The temporal lobe response on the affected side is decreased [22].

3.2 Functional Recovery, Plasticity, and Indices of Reorganization

  1. 1.

    Spontaneous magnetic fields (slow waves)

    • Slow waves on the affected side decrease with improved circulation after surgical treatment [6, 23].

    • Normalization of slow waves leads to a clinical improvement in symptoms [12].

    • Revascularization improves CBF and decreases theta bands, but there are no changes in delta bands [23]. In another report, delta waves near the lesion did not change during follow-up, and there was no correlation with clinical symptoms [3].

    • In TIAs, normalization of slow waves from the somatosensory cortex may be an index of short-term functional recovery after stroke [24].

    • In cases of aphasia due to stroke, delta waves (1–4 Hz) were seen near the lesion, and a decrease of delta waves after speech therapy was associated with good outcomes [25].

    • Delta waves on the unaffected side and gamma waves on the affected side may be indices of functional recovery.

  2. 2.

    Evoked magnetic fields

    • Improved SEF latency correlates with sensory improvement [14].

    • Source power localized to S1 (SEF) and to M1 with finger tapping showed a correlation with sensory and motor function improvement [26].

4 Results 2

Of the abovementioned 28 papers, 14 with high evidence levels, including measurement of CBF or comparisons with normal controls, were further reviewed. Table 11.1 describes their content in detail.

Table 11.1 Summary of some meritorious manuscripts
Full size table

4.1 Target Diseases and Comparisons

Most papers discussed internal carotid artery or middle cerebral artery occlusion or stenosis. One paper dealt with TIAs, and one paper was about moyamoya disease. There were normal controls in 11 papers.

4.2 Analysis Methods

Spontaneous magnetic fields were measured in six studies, evoked cerebral magnetic fields were measured in six studies, and both spontaneous and evoked cerebral magnetic fields were measured in one study. Among papers dealing with evoked cerebral magnetic fields, four used median nerve stimulation, one used tactile finger stimulation [15], one used auditory stimulation [4], and two used motor-related magnetic fields [27, 28].

Most papers on evoked magnetic fields used a single equivalent current dipole (ECD) method, but one used a spatial filtering technique (synthetic aperture magnetometry, SAM) [28]. Measurement of spontaneous magnetic fields included analysis of the waveforms themselves, analysis of magnetic field distribution using ECDs, analysis using a spatial filtering technique (standardized low-resolution brain electromagnetic tomography, sLORETA) [23], and analysis using power spectral density (PSD) [4].

One paper showed the relation between direct current (DC)-MEG signal and near-infrared spectroscopy (NIRS) for a finger movement task [27].

4.3 Cerebrovascular Measurements

Six papers described measurement of CBF, including five that measured cerebrovascular reserve capacity. Three papers discussed areas of decreased cerebrovascular reserve (penumbra). They mainly described the correlation with areas of slow wave appearance and discussed postoperative improvement. Ohtomo and coworkers [6] demonstrated typical temporoparietal theta activity (TPTA) in ischemic lesion in Fig. 11.1 and the relation between cerebral circulation and TPHA in Fig. 11.2. Sakamoto and coworkers [23] showed the changes in slow wave activity between pre- and postsurgical anastomosis after stroke in Fig. 11.3.

Fig. 11.1
figure 1

A 54-year-old male with temporoparietal theta activity (TPTA) detected by MEG. (a) Digital subtraction angiogram showing stenosis (arrow) of the right internal carotid artery. (b) T2-weighted MR image showing a right striatocapsular infarct lesion. (c) SPECT scan showing flow reduction in the territory of the right middle cerebral artery. (d) MEG waveforms, detected in the awake condition with the eyes closed using the latitudinal and longitudinal tangential derivatives of 204 planar-type gradiometers, showing TPTA in the right hemisphere. (e) Isofield map at a typical peak of TPTA (broken line in the inlet square of D), showing a single dipole pattern. The arrow shows the approximate location and orientation of the equivalent current dipole (ECD) of TPTA over the right temporal area. (f) ECDs at ten similar peaks of TPTA to (d) and (e) projected onto three orthogonal MR images. Circles and bars indicate the ECD location and orientation, respectively (Ohtomo et al. [15])

Full size image
Fig. 11.2
figure 2

Resting regional cerebral blood flow (rCBF) and regional cerebrovascular reactivity (rCVR) in the middle cerebral artery territory of 112 hemispheres in 56 patients with or without temporoparietal theta activity (TPTA). Dashed vertical and horizontal lines indicate the criteria for reduced rCBF (43.4 ml/100 g brain/min) and reduced rCVR (10 %) as defined in normal subjects in our institute. Note that TPTA was detected in six of nine hemispheres in the patients with both reduced rCBF and reduced rCVR (Ohtomo et al. [15])

Full size image
Fig. 11.3
figure 3

(a) A 70-year-old man with the severe stenosis of right internal carotid artery showed slow wave activity in preoperative state by sLORETA of MEG. (A) MRI (FLAIR) showed no severe cerebral infarction, excluding old infarct lesion in the left parietooccipital area (arrowhead) and multiple lacunar infarctions in bilateral basal ganglia and periventricular white matter. (B) Cerebral angiography showed N90 % stenosis of the C5 segment of the right internal carotid artery (arrow). (C, D) 15O-gas PET showed reduced CBF (C, 32.5 ml/100 g/min) and increased OEF (D, 73.3 %) in the right cerebral cortex. (E) Quantitative imaging of cerebral neuromagnetic fields using sLORETA-qm in four frequency bands showed distribution and increased intensities of slow waves dominantly in the parietooccipital area of the right cerebral hemisphere. Distribution of slow waves was also recognized in the left parietooccipital area and periventricular white matter corresponding with old infarct lesion (Sakamoto et al. [19]). (b) Postoperative investigations in same case. (A) MRI (DWI) showed no additional infarction in any area of cerebrum. (B) Cerebral angiography showed recovery of stenosis after CAS of the right internal carotid artery (arrow). (C, D) 15O-gas PET showed increased CBF (C, 38.0 ml/100 g/min) and decreased OEF (D, 47.1 %) in the right cerebral cortex. (E) Quantitative imaging of cerebral neuromagnetic fields with sLORETA-qm in each frequency band showed decreased intensities at 2–4 Hz, 4–6 Hz, and 6–8 Hz in the parietooccipital area of the right cerebral hemisphere, although no apparent change of intensity at 0.3–2 Hz was observed compared with preoperative data (Sakamoto et al. [19])

Full size image

5 General Remarks

Low-frequency activity often appears near ischemic foci and in an ischemic hemisphere, and these waves also tend to occur in the temporoparietal area [5]. In addition, with improved cerebral blood flow (CBF), slow waves, particularly theta band activity, usually disappear [6, 23]. However, this phenomenon is not necessarily seen in all cases, so using changes in slow waves as an objective index of a penumbra can be unreliable.

Many investigators have reported a decreased SEF N20m response to median nerve stimulation in an ischemic hemisphere, but it is difficult to determine whether this signifies complete ischemia or a penumbra.

MEG reflects activity in the sulcal cortex, whereas EEG also reflects activity in the sulcal and gyral cortex activity. On the other hand, MEG uses large number of channels (> 100 sensors on the whole head), so MEG can detect brain function over a wider area. Furthermore, its spatial resolution is much better than that of EEG. MEG directly detects magnetic fields because it is unaffected by the skin and bones and is, therefore, quantitatively superior to EEG. Moreover, MEG and MRI images can be fused, making it convenient for clinical application, such as neurosurgical navigation techniques.

MEG is a noninvasive technique to quantify neural activity, but it is not yet an established method for evaluating cerebral ischemia. The use of new analytical methods, including spatial filtering, not only measuring spontaneous magnetic fields but also objectively assessing neural activity when patients perform simple tasks [28], is a novel approach using cerebral magnetic fields to evaluate ischemic cerebrovascular disease.

6 Summary

  1. 1.

    Slow waves appear on the affected side in patients with stroke.

  2. 2.

    Slow waves are seen in areas of decreased blood perfusion (often the temporoparietal area) in patients with cerebrovascular occlusion.

  3. 3.

    The first peak of the somatosensory evoked field (SEF) in the ischemic hemisphere may decrease, may disappear, or be abnormal.

7 Conclusion

This section examined trends in clinical research and reviewed the scientific evidence, based on a literature search, for the use of MEG in evaluating ischemic cerebrovascular disease. MEG could not absolutely evaluate brain plasticity and circulation. Although it is complete as an alternative device to other modalities, a few papers reported a relatively high level of evidence for using MEG in diagnosis and treatment planning. MEG may be superior to scalp EEG in assessing ischemic changes in neural function.

MEG may also be useful to objectively evaluate neurological function, other than from the perspective of cerebral blood flow and metabolism, in patients with ischemic cerebrovascular disease.