Mild traumatic brain injury (MTBI) is a major public health concern because of its high incidence, which has been reported to account for at least 75% of all brain injuries (Sosin et al., 1996). Postconcussion syndrome (PCS) occurs after MTBI and includes a number of disparate symptoms such as headache, dizziness, fatigue, irritability, difficulty in concentration and performing mental tasks, cognitive impairments, and reduced tolerance to stress. Generally, patients with minor concussive injuries recover within several days; however, approximately 15% of patients in whom MTBI has been diagnosed may present with persistent, debilitating problems (Alexander, 1995; Kushner, 2002). Documented abnormalities after mild head injury that were reported in neuropathologic (King, 1997; Pearl, 1998), neurophysiologic (Gaetz and Weinberg, 2000; Watson et al., 1995), neuroimaging (Kant et al., 1997; Kushi et al., 1994), and neuropsychological (Jacobson, 1995; Trudeau et al., 1998) studies suggest that the syndrome has an organic background, though its pathophysiology is not yet known. Moreover, whereas persistent symptoms and cognitive deficits are present in a minority of PCS patients for several months to years, the significant risk factors and underlying pathogenesis for persisting sequelae remain controversial (De Kruijk et al., 2002).
In this study, we retrospectively investigated EEG recordings and brain images of 17 PCS patients 1 month to 7 years after their respective head injuries. To further explore the possible pathogenesis, we quantitatively evaluated their digitized EEG data and used low-resolution brain electromagnetic tomography (LORETA) to localize the sources of abnormal activity in comparison with respective imaging studies. We show that focal cortical slowing is a characteristic finding in PCS patients and that it may be associated with persistent, increased blood–brain barrier (BBB) permeability and regional blood flow (rCBF) deficit.
The study protocol was approved by the local institutional medical ethics board and Helsinki Committee. All patients were chosen from our routine follow-up consult database in the neurosurgical outpatient clinic of our medical center. The inclusion criteria for patients entering this study were (1) a history of mild head trauma (Glasgow Coma Scale score at admission > 12); and (2) a display of persistent symptoms consistent with ICD-10 PCS criteria more than 1 month after the injury. Seventeen PCS patients (12 males and 5 females) aged 17 to 49 (29.9 ± 11.2) years who were examined at our neurosurgical outpatient clinic between the years 2000 to 2002 were included in this study (Table 1). The control group (10 males and 7 females) included healthy volunteers with no history of significant brain injury or neurologic disorders and with a mean age similar to the patients (32.1 ± 10.7 years, P > 0.5).
Sixteen of the 17 patients were admitted to the trauma center with a Glasgow Coma Scale score of 15 and a diagnosis of isolated mild head trauma; 10 had a significant loss of consciousness at the time of trauma. This was the typical indication for admission in the presence of normal brain computed tomography (CT) findings (n = 7). Three patients had small intracerebral hemorrhagic contusions on the day of admission, with no mass effect or focal neurologic deficit. In all three, surgical intervention was not indicated and complete hemorrhage resolution was observed when brain CT was repeated during hospitalization (<7 days). Repeated brain CT scans 3 to 6 weeks after the trauma were normal in all cases. Five patients were examined on the day of the trauma at the emergency room and discharged after normal neurologic examination and normal brain CT. One patient (#5 in Table 1) was referred for neurosurgical consultation 3 days after the trauma because of severe headache. He was discharged for further follow-up after normal neurologic examination and no pahological CT findings. Another patient (#4) was admitted to the emergency room with a Glasgow Coma Scale score of 8 due to subtentorial acute epidural hematoma and a brain stem compression. This patient regained full consciousness shortly after surgical evacuation of the hematoma and withdrawal of anesthetics (<12 hours after admission). This patient was the only one in this series with multiple injuries (pneumothorax, spleen and liver lacerations), which were the indicators for his prolonged hospitalization (18 days). Two patients had linear skull fractures. Patients were reevaluated at the neurosurgical outpatient clinic between 1 month and 7 years (17 ± 22 months) after the trauma. Further investigative studies were recommended for patients who continued to display significant and persistent symptoms more than 4 to 8 weeks after the trauma. The most frequent reported symptoms were headache, fatigue and decreased cognitive capabilities. In all cases a complete neurologic examination was documented as normal at the time of follow-up. In addition to QEEG and brain CT (n = 17), patients were subjected to brain imaging studies including brain MRI (n = 3), rCBF 99mTc-ethylcysteinate dimer (Tc-ECD) single photon emission computed tomography (SPECT) (n = 13), and 99mTc-diethylenetriaminepentaacetic acid (Tc-DTPA) SPECT (n = 11) scans.
Recordings were made using a 128-channel digital EEG acquisition unit (CEEGRAPH IV, Bio-logic Systems Corp., Mundelein, IL, U.S.A.). All patients and subjects underwent quantitative EEG (QEEG) either with an elastic 128-aluminum electrode–embedded cap with conductive gel (Electro-Cap International, Inc., Eaton, OH, U.S.A.) with preset electrode positions within the standard international 10–20 positioning system (14 and 11 out of 17 from the control and patients groups respectively), or with 23 individual aluminum cup electrodes with conductive paste (Tomkins et al., 2001). Analysis was restricted to data from electrodes displaying impedances of less than 10 KΩ. EEG data were digitally filtered online using a bandwidth of 1 to 30 Hz, and periods of artifactual data were noted online. A minimum of 20 minutes of clean EEG data were recorded from each patient and control.
Localization of electrical neuronal activity was computed with the LORETA inverse solution (Pascual-Marqui et al., 1994, 1999). This procedure computes, from the recorded scalp electrical potential differences, the three-dimensional distribution of the electrically active neuronal generators in the brain as a current density value (A/m2) at each voxel. The LORETA linear solution is characterized by the property that the activity at any given voxel in cortical gray matter must be as similar as possible to the average activity of its neighboring voxels. This property corresponds to the mathematical implementation of synchronized activity among neighboring neurons (Llinas, 1988; Sukov and Barth, 1998)—a necessary condition for the generation of the EEG (reviewed by Dale et al., 2000; Pascual-Marqui et al., 2002). It has been shown in several simulation experiments that the method is capable of correct localization with fairly low errors (Menendez et al., 2001; Pascual-Marqui, 1999; Pascual-Marqui et al., 2002) and (Phillips et al., 2002). The version of LORETA used here used a three-shell spherical head model registered to the digitized atlas Talairach and Tournoux, 1988; Brain Imaging Centre, Montreal Neurologic Institute). Electrode coordinates were calculated by cross-registration between spherical and realistic head geometry (Towle et al., 1993). Computations were restricted to cortical gray matter and hippocampi according to the digitized Probability Atlas (Brain Imaging Centre, Montreal Neurologic Institute). A voxel was labeled as gray matter if its probability of being gray matter (1) exceeded 33%, (2) exceeded the probability of being white matter, and (3) exceeded the probability of being cerebrospinal fluid. In the current implementation, a spatial resolution of 7 mm was used; the result is that for each instantaneous scalp potential distribution map, there is a three-dimensional LORETA image consisting of 2,394 voxels in total. Pascual-Marqui et al. (2002) reviewed the experimental validation for LORETA based on the correct localization of primary sensory cortices, epileptic foci, language processing areas, and face processing areas.
The computation of electrical neuronal activity for any arbitrary EEG frequency band is based on the principle that the activity (power) for a given band corresponds to the average activity (power), over time, obtained from the band-filtered EEG (Pascual-Marqui et al., 1999a). Therefore, specific frequency band activity is given by the time-average LORETA image for the filtered EEG data, where each instantaneous LORETA image consists of squared current density amplitudes (power) at each voxel. However, identical results are obtained more efficiently in the frequency domain, using EEG cross-spectra, due to the equivalence of power in the time and in the frequency domains (see, for example, Cooley et al., 1977). A thorough account containing all the technical details of these computations can be found elsewhere (Frei et al., 2001).
Linear inverse solutions such as LORETA are known to be sensitive to the presence of noise in the measurements. A partial solution to this problem is to “oversmooth” or regularize the solution, which then, consequently, will not explain the measurements exactly because they are likely to be contaminated with noise. In this study, the amount of regularization was determined by cross-validation, which is an automatic and objective method that optimizes the predictive power of the solution (Pascual-Marqui, 1999a; Pascual-Marqui et al., 1999b).
Single Photon Emission Computed Tomography
All SPECT scans were performed in patients of the study population as part of their clinical follow-up and investigation. In all cases, SPECT was done within 4 weeks after QEEG. SPECT was done using a dual headed “Varicam” gamma camera connected to an image processing “Expert” computer (acquisition mode: 128 × 128-pixel matrix; 120 images, each image at 3 degrees). rCBF was measured after the administration of the freely permeable compound Tc-ECD as the radioactive tracer. BBB integrity was evaluated after the administration of the nonpermeable compound Tc-DTPA. Tc-DTPA SPECT scans were done 1 to 2 weeks after perfusion scans.
Data Analyses and Statistics
EEG data were visually inspected and 60 to 80 seconds of artifact-free, closed-eye EEG data were extracted for quantitative analysis. Fast Fourier transform was applied to the EEG waveforms recorded from each electrode of each subject. The average discrete frequency was normalized to each subject's own total power of the 1.5- to 30-Hz frequency spectrum. Data are expressed throughout the text as means and standard errors of the mean (SEMs). Statistically significant differences between control and patient LORETA values where tested by using the post hoc t-test, on a voxel-by-voxel basis (Strik et al., 1998). Other significant differences between groups were determined by Student's t-tests, with a P value of less than 0.05 as the level of statistical significance. Regarding imaging, in cases of overlapping hypoperfusion and compromised BBB, the respective Tc-ECD and Tc-DTPA tomographic images were overlaid for colocalization of findings.
QEEG Revealed Abnormal Slowing in PCS Patients
Postconcussion patients showed a significant increase in power within the delta band (1.5–5 Hz), and a decrease in power within the alpha band (8.5–12 Hz) (Fig. 1A). For each patient's normalized power spectrum, the distance in number of SEMs (of control) was calculated and plotted (Fig. 1B). In 16 out of 17 patients, clear peaks of abnormal power were found within the delta (1.5–4.5 Hz, n = 11) or the delta-theta (6–7 Hz, n = 5) ranges. In addition, seven patients also displayed increased power within the beta-1 (12.5–18 Hz), and one in the alpha-2 ranges (10.5–12 Hz, n = 1). In our series, only one patient had an abnormal increase power only at the beta range with no significant increase at slower bands (Fig 1B).
LORETA was used to further explore the source for abnormal rhythms in three-dimensional space. Figure 2 A shows LORETA values (correlating to current density) for the delta band generators in each of the 2,394 voxels representing cortical gray matter (inset). Whereas controls displayed generators of delta rhythm in consistent voxels representing distinct cortical regions, PCS patients displayed increased intervoxel variability for these generators, where voxels of maximal activity varied from patient to patient (Fig. 2B). The averaged source of the delta in the patient group was equally localized in peripheral neocortical regions of all lobes (Fig. 2C). This was in contrast to the centrally located source along the cingulate gyrus in healthy controls (figure is two dimensional). Statistical comparison revealed significant differences between the two groups in both the delta and alpha frequency ranges, with significant excess activity located in peripheral regions (red voxels in Fig. 2E).
LORETA Localizes Abnormal Slowing to Regions of Reduced Perfusion and Abnormal BBB
The large variability in the source for abnormal slow activity in our patient group suggests that it is unlikely to originate from one common source, rather from various cortical regions, possibly related to the trauma. In five patients, maximal sources for abnormal activity were found to be in medial cingulate gyrus, similar to that found in controls. In the other 12 patients, maximal activity involved the left frontal (Brodman's area: B-8,10; n = 7 patients), left temporal (B-39; n = 4), right frontal (B-8,10; n = 4; bilateral sources were found in two patients), right parietal (B-5,7; n = 2), and right occipital (B-18, n = 1) areas (Table 1). There was a tendency for higher number of patients with abnormal cortical sources in the left hemisphere compared with the right one (seven versus two patients; three showed bilateral sources), but this different was not found to be statistically significant (Pearson χ2). Notably, all three patients with a focal parietal slowing reported events of paraesthesia in the contralateral limb (see Discussion).
For each patient, the maximal source for the abnormal rhythm was determined and compared with SPECT findings (Table 1). In eight patients, LORETA localized the source of the abnormal activity at or closely related (at the same cortical lobe) to the pathologic region observed in SPECT. Figure 3 shows imaging, power spectrum, and LORETA analyses from two patients with a persistent PCS more than 3 years after the trauma compared with averaged control (Fig. 3A). Patient 1 (#13 in Table 1) was a 41-year-old healthy man who was involved in an automobile accident and continued to suffer from headaches, general weakness, and occasional losses of consciousness without aurae. Behavioral examination documented slow task performance with decreased attention, concentration, and verbal memory as well as dyscalculia. The patient was referred to our laboratory 7 years after trauma. QEEG revealed increased power at the delta and beta-1 bands. LORETA localized both abnormal rhythm generators in the left inferior frontal gyri (B-9,45; Fig. 3B). rCBF-SPECT showed hypoperfused foci in the left medial frontotemporal regions colocalizing with BBB disruption showed by Tc-DTPA SPECT. Patient 2 (#17), a 49-year-old man, was involved in an automobile accident resulting in MTBI. He continued to suffer from persistent PCS 3.5 years after trauma and was referred for a QEEG, which revealed increased power at the theta and beta-1 bands as well as a prominent decrease in the alpha band. For all three abnormal bands, LORETA identified a single prominent source in the left middle frontal gyrus (B-9) that colocalized with both rCBF decrease and abnormal BBB (Fig. 3C).
Resolution of QEEG and Imaging Findings
Figure 4 shows imaging and QEEG findings in one patient studied at both 6 weeks and 10 months after trauma (#8 in Table 1). In this patient, EEG data collected in the first visit showed occasional interictal like spikes. LORETA localization of the maximal current density of an average of 13 spikes indicated the right parietal region (B-2) as the underlying generator. The same region showed a concurrent decrease in Tc-ECD enhancement and an increase in Tc-DTPA signal enhancement, respectively, indicating focal rCBF decrease and a compromised BBB. Power-frequency spectrum analysis of his averaged EEG showed abnormal peak at 6.5 Hz. This abnormal activity localized to the bilateral parietal lobes, although more prominently in the right. The patient was treated with antiepileptics (carbamazepine), which were discontinued after 4 months of gradual improvement. Ten months after the trauma, after complete clinical resolution, EEG and SPECT were repeated. Both Tc-ECD and Tc-DTPA SPECT images revealed no pathologic findings. Likewise, closed-eye QEEG analyses showed normalization of both the power spectrum and the LORETA localization of the maximal source of the respective frequency bands (Fig. 4D).
Our combined physiologic and imaging findings from our PCS patient group support several conclusions. First, in most PCS patients, QEEG power spectrum analysis reveals significant abnormalities, most often an increase in slow wave delta together with a decrease in alpha activity. Second, abnormal QEEG represents a focal cortical generator that varies in location rather than a diffuse pathology. Third, abnormal focal rhythmogenesis may be associated with a long-lasting focal increase in BBB permeability and decrease in rCBF. Finally, both the neurophysiologic and BBB abnormalities may last for years in persistent PCS or be reversible in the more transient form of the syndrome.
EEG Changes in Patients With PCS
EEG is still the most available method for recording brain electrical activity in humans. Abnormal spontaneous (Watson et al., 1995) and evoked (i.e., event related potentials) (Gaetz and Weinberg, 2000) scalp recordings have been previously described and provided a strong basis for the current view that PCS is an organic syndrome. Several authors have described an increase in theta power and a decrease in alpha as the most obvious measurable quantitative changes in PCS patients EEG data (Shaw, 2002). However, the sources of the abnormal activity, its relation to clinical syndrome, and its time dependency or underlying mechanisms have not been described. Similar to the aforementioned published findings, the results from our series of PCS patients show a significant decrease in the alpha band and a significant overall increase in the delta and theta bands, although in some patients a significant increase was also found in the beta band. LORETA was used to further explore the underlying generators for each classical frequency band. On average, the PCS patients' abnormal rhythms were found to originate in peripheral cortical regions compared with the midline-restricted symmetric source characterizing our control group (Fig. 2). Moreover, when comparing LORETA values for each specific brain voxel, a wide variability between patients was found, in contrast to the relatively low variability among the control group. This variability in LORETA values presumably represents variable, focal sources for respective abnormal generators. These findings suggest that in PCS patients, focal cortical abnormal generators that differ in location between patients are related to the location of cerebral injury rather than to a single common “pathologic” generator or a general diffused cortical slowing, as one might expect in case of a general stress-dependent mechanism or lesion in deeper brain structures like the thalamus or brain stem (Andy, 1989; Soustiel et al., 1995).
The functional significance of the abnormal EEG and its relation to patient symptomatology and prognosis is an interesting debate. In a prospective study, Watson et al. (1995) followed up 26 patients after minor closed head injury. He found 50% to have residual symptoms after 1 year, with a significant theta/alpha ratio increase and correlation between EEG and symptom counts at 6 weeks. Moreover, in their study there was a direct relationship between the patient's recovery time and the extent of his or her symptomatology, suggesting an association between abnormal cortical activity, symptomatology and the course of the respective disease. In our series, several lines of evidence support the view that cortical dysfunction, as reflected by EEG findings, may underlie PCS. First, three patients who also presented with transient attacks of paraesthesia, clinically suggestive for partial epilepsy, had a focal abnormal EEG localized to contralateral parietal cortex. Second, the correlation between EEG aberrations and reduced rCBF perfusion and BBB permeability suggests an association between the functional and anatomic lesions (see below). Third, in eight patients with persistent PCS (including two cases of more than 3 years' duration), EEG abnormalities were related anatomically to the SPECT findings, whereas in one patient in whom the clinical syndrome was resolved, recovery was associated with parallel resolution of both EEG aberrations and SPECT findings. Thus, our study supports the possibility that in at least some PCS patients, focal cortical vascular lesions are associated with BBB disturbances and reduced blood flow as the underlying pathogenesis. It is not yet clear if and how a focal cortical dysfunction leads to the general symptoms that characterize many PCS patients. One likely mechanism could be associated with disturbances in corticothalamic or cortico-cortico connectivity, as was shown using altered coherence between EEG electrodes in PCS patients (Thatcher et al., 1989).
The Anatomic Lesion
Our data suggest that a vascular lesion and a specifically compromised BBB as the possible anatomic basis for PCS. Although only three patients underwent brain MRI to exclude a pathology for which baseline CT is insensitive (e.g., white matter lesion), our results are consistent with previous studies showing that routine anatomic brain imaging techniques like CT and MRI do not show apparent abnormality in the majority of PCS patients (Ingebrigtsen et al., 1998). Similar to our study, Kant et al. (1997) found Tc-ECD SPECT (positive in 53% of PCS patients) to be more sensitive than MRI (9%) or CT (4.6%) for the detection of PCS-related brain pathology (Kant et al., 1997). In 11 of the 13 patients for whom ECD-SPECT was available, a focal reduction in cortical perfusion was evident. Interestingly, DTPA-SPECT scans showed regions of abnormal enhancement, indicating a compromised BBB. The high sensitivity of SPECT-DTPA in identifying BBB disruption is probably due to its high sensitivity in detection injected agents (picomolar range) compared with MRI (micromolar range) (Volkow et al., 1997).
Blood–brain barrier disruption is a well-known pathologic finding in both animal studies (Shapira et al., 1993) and humans after acute head trauma (Cervos-Navarro and Lafuente, 1991). However, to the best of our knowledge, it was never shown to persist for weeks to years, and likewise the functional significance of such a lesion lies unknown. In 28 patients with severe traumatic brain injury, Csuka et al. (1999) tracked BBB permeability by measuring albumin CSF-serum levels for 22 days after trauma. They found mild to severe BBB disruption in 19 patients. Generally, disruption appeared within 24 hours of the trauma. Restoration to normal BBB was achieved within 8 to 19 days after injury but was documented for only one third of the patients included in that study. Kushi (1994) showed increased cerebrovascular permeability to gadolinium-DTPA in areas around cerebral contusions beginning 1 to 2 days after the traumatic incident. Although we did not evaluate BBB permeability immediately after the injury, our data do suggest that BBB disruption may last for weeks to years after the injury.
Other important unknowns are the consequences of BBB disruption. Although in the work of Csuka (1999) patients with severe BBB disruption had worse prognoses, there are no data presented to normalize for trauma severity. Inducing BBB disruption in the brain cortices of animal models by direct bile salt cortical application results in the delayed appearance of focal cortical epileptiform activity (Seiffert et al., 2004; Zappulla et al., 1985). Although the detailed mechanisms underlying this activity are not known, it is arguable that such changes in cortical function are expected after BBB disruption; increased cerebrovascular permeability not only causes edema and reduces perfusion by inducing a shift of proteins and water into the extracellular brain environment, but also tends to equilibrate ionic concentrations between serum and CSF, leading to, for example, increased K+ and decreased Ca2+ and Mg2+ concentrations in the neuronal microenvironment. Similarly, glutamate and other excitatory amino acids that are of higher concentration in the serum than in the CSF may shift. Numerous studies on isolated brain slices in vitro show that such changes result in increased cortical excitability, hypersynchronicity, and “epileptiform”-like activity (Andrew, 1991; Rosen and Andrew, 1990; Saly and Andrew, 1993; Schwartzkroin et al., 1998). Indeed, concomitant BBB disruption and seizures are well described in epilepsy (Oztas and Turkel, 2001), although there are only few direct experimental data that probe the relationship between the two phenomenons. Slow, epileptiform activity is also a typical finding in animal models of mild head trauma (Nilsson et al., 1994), in the injured cortex (Prince and Tseng, 1993), and in several reports involving humans (Herman, 2002; McCrory and Berkovic, 2000). Interestingly, strong experimental evidence suggests the concussion is related to cortical seizure activity (Walker et al., 2002); however, this is the first report to suggest that the lasting BBB disruption may follow brain injury and correlate with focal cortical dysfunction and PCS symptoms.
Alexander MP. (1995) Mild traumatic brain injury: pathophysiology, natural history, and clinical management. Neurology
Andrew RD. (1991) Seizure and acute osmotic change: clinical and neurophysiological aspects. J Neurol Sci
Andy OJ. (1989) Post concussion syndrome. Brainstem seizures: a case report. Clin Electroencephalogr
Cervos-Navarro J, Lafuente JV. (1991) Traumatic brain injuries: structural changes. J Neurol Sci
Cooley JW, Lewis PAW, Welch PD. (1977) The fast Fourier transform and its application to time series analyses. In: Enslein K, Ralston A, Wilf HS, eds. Statistical methods for digital computers
. New York: Wiley, 377–423.
Csuka E, Morganti-Kossmann MC, Lenzlinger PM, Joller H, Trentz O, Kossmann T. (1999) IL-10 levels in cerebrospinal fluid and serum of patients with severe traumatic brain injury: relationship to IL-6, TNF-alpha, TGF-beta1 and blood-brain barrier function. J Neuroimmunol
Dale AM, Liu AK, Fischl BR, Buckner RL, Belliveau JW, Lewine JD, Halgren E. (2000) Dynamic statistical parametric mapping: combining fMRI and MEG for high-resolution imaging of cortical activity. Neuron
De Kruijk JR, Leffers P, Menheere PP, Meerhoff S, Rutten J, Twijnstra A. (2002) Prediction of post-traumatic complaints after mild traumatic brain injury: early symptoms and biochemical markers. J Neurol Neurosurg Psychiatry
Frei E, Gamma A, Pascual-Marqui R, Lehmann D, Hell D, Vollenweider FX. (2001) Localization of MDMA-induced brain activity in healthy volunteers using low resolution brain electromagnetic tomography (LORETA). Human Brain Mapp
Gaetz M, Weinberg H. (2000) Electrophysiological indices of persistent post-concussion symptoms. Brain Inj
Herman ST. (2002) Epilepsy after brain insult: targeting epileptogenesis. Neurology
Ingebrigtsen T, Waterloo K, Marup-Jensen S, Attner E, Romner B. (1998) Quantification of post-concussion symptoms 3 months after minor head injury in 100 consecutive patients. J Neurol
Jacobson RR. (1995) The post-concussional syndrome: physiogenesis, psychogenesis and malingering. An integrative model. J Psychosom Res
Kant R, Smith-Seemiller L, Isaac G, Duffy J. (1997) Tc-HMPAO SPECT
in persistent post-concussion syndrome after mild head injury: comparison with MRI/CT. Brain Inj
King N. (1997) Mild head injury: neuropathology, sequelae, measurement and recovery. Br J Clin Psychol
Kushi H, Katayama Y, Shibuya T, Tsubokawa T, Kuroha T. (1994) Gadolinium DTPA-enhanced magnetic resonance imaging of cerebral contusions. Acta Neurochir Suppl (Wien)
Kushner D. (2002) Toward an evidence-based approach in the management of concussion: the role of neuroimaging. AJNR Am J Neuroradiol
Llinas RR. (1988) The intrinsic electrophysiological properties of mammalian neurons: insights into central nervous system function. Science
McCrory PR, Berkovic SF. (2000) Video analysis of acute motor and convulsive manifestations in sport-related concussion. Neurology
Menendez RGD, Andino SG, Lantz G, Michel CM, Landis T. (2001) Noninvasive localization of electromagnetic epileptic activity. I. Method descriptions and simulations. Brain Topogr
Nilsson P, Ronne-Engstrom E, Flink R, Ungerstedt U, Carlson H, Hillered L. (1994) Epileptic seizure activity in the acute phase following cortical impact trauma in rat. Brain Res
Oztas B, Turkel N. (2001) Influence of an abrupt increase in blood pressure on the blood-brain barrier permeability during acute hypertension and epileptic seizures. Pharmacol Res
Pascual-Marqui RD. (1999a) Reply to comments made by R. Grave De Peralta Menendez and S.I. Gozalez Andino. Int J Bioelectromag
1(2). Available at: http://www.ee.tut.fi/rgi/ijbem/volume1/number2/html/pascual.htm.
Pascual-Marqui RD, Esslen M, Kochi K, Lehmann D. (2002) Functional imaging with low-resolution brain electromagnetic tomography (LORETA): a review. Methods Find Exp Clin Pharmacol
24:suppl C 91–5.
Pascual-Marqui RD, Lehmann D, Koenig T, Kochi K, Merlo MC, Hell D, Koukkou M. (1999b) Low resolution brain electromagnetic tomography (LORETA) functional imaging in acute, neuroleptic-naive, first-episode, productive schizophrenia. Psychiatry Res
Pascual-Marqui RD, Michel CM, Lehmann D. (1994) Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain. Int J Psychophysiol
Pearl GS. (1998) Traumatic neuropathology. Clin Lab Med
Phillips C, Rugg MD, Friston KJ. (2002) Anatomically informed basis functions for EEG source localization: combining functional and anatomical constraints. Neuroimage
Prince DA, Tseng GF. (1993) Epileptogenesis in chronically injured cortex: in vitro studies. J Neurophysiol
Rosen AS, Andrew RD. (1990) Osmotic effects upon excitability in rat neocortical slices. Neuroscience
Saly V, Andrew RD. (1993) CA3 neuron excitation and epileptiform discharge are sensitive to osmolality. J Neurophysiol
Schwartzkroin PA, Baraban SC, Hochman DW. (1998) Osmolarity, ionic flux, and changes in brain excitability. Epilepsy Res
Seiffert E, Dreier JP, Ivens S, Bechmann I, Heinemann U, Friedman A. (2004) Focal blood-brain barrier disruption induces epileptic focus in the rat somatosensory cortex. J Neurosci
Shapira Y, Setton D, Artru AA, Shohami E. (1993) Blood-brain barrier permeability, cerebral edema, and neurologic function after closed head injury in rats. Anesth Analg
Shaw NA. (2002) The neurophysiology of concussion. Prog Neurobiol
Sosin DM, Sniezek JE, Thurman DJ. (1996) Incidence of mild and moderate brain injury in the United States, 1991. Brain Inj
Soustiel JF, Hafner H, Chistyakov AV, Barzilai A, Feinsod M. (1995) Trigeminal and auditory evoked responses in minor head injuries and post-concussion syndrome. Brain Inj
Strik WK, Fallgatter AJ, Brandeis D, Pascual-Marqui RD. (1998) Three-dimensional tomography of event-related potentials during response inhibition: evidence for phasic frontal lobe activation. Electroencephalogr Clin Neurophysiol
Sukov W, Barth DS. (1998) Three-dimensional analysis of spontaneous and thalamically evoked gamma oscillations in auditory cortex. J Neurophysiol
Talairach J, Tournoux P, eds. (1988) Co-planar stereotaxic atlas of the human brain: three-dimensional proportional system
. Stuttgart: Georg Thieme.
Thatcher RW, Walker RA, Gerson I, Geisler FH. (1989) EEG discriminant analyses of mild head trauma. Electroencephalogr Clin Neurophysiol
Tomkins O, Kaufer D, Korn A, Shelef I, Golan H, Reichenthal E, Soreq H, Friedman A. (2001) Frequent blood-brain barrier disruption in the human cerebral cortex. Cell Mol Neurobiol
Towle VL, Bolanos J, Suarez D, Tan K, Grzeszczuk R, Levin DN, Cakmur R, Frank SA, Spire JP. (1993) The spatial location of EEG electrodes: locating the best-fitting sphere relative to cortical anatomy. Electroencephalogr Clin Neurophysiol
Trudeau DL, Anderson J, Hansen LM, Shagalov DN, Schmoller J, Nugent S, Barton S. (1998) Findings of mild traumatic brain injury in combat veterans with PTSD and a history of blast concussion. J Neuropsychiatry Clin Neurosci
Volkow ND, Rosen B, Farde L. (1997) Imaging the living human brain: magnetic resonance imaging and positron emission tomography. Proc Natl Acad Sci U S A
Walker MC, White HS, Sander JW. (2002) Disease modification in partial epilepsy. Brain
Watson MR, Fenton GW, McClelland RJ, Lumsden J, Headley M, Rutherford WH. (1995) The post-concussional state: neurophysiological aspects. Br J Psychiatry
Zappulla RA, Spigelman MK, Omsberg E, Rosen JJ, Malis LI, Holland JF. (1985) Electroencephalographic consequences of sodium dehydrocholate-induced blood-brain barrier disruption: Part 1. Acute and chronic effects of intracarotid sodium dehydrocholate. Neurosurgery