PEDIATRICS Vol. 106 No. 3 September 2000, pp. 527-532

Neuroimaging in Children With Newly Diagnosed Epilepsy: A Community-Based Study

Anne T. Berg, PhD^*, Francine M. Testa, MD, Susan R. Levy, MD, and Shlomo Shinnar, MD, PhD^§

From ^* Northern Illinois University Department of Biological Sciences, DeKalb, Illinois;  Yale University Departments of Pediatrics and Neurology, New Haven, Connecticut; and ^§ Montefiore Medical Center, Albert Einstein College of Medicine Departments of Neurology and Pediatrics and the Comprehensive Epilepsy Center, Bronx, New York.

	ABSTRACT

Top Abstract Methods Results Discussion References

Background.  Neuroimaging is generally considered a part of the evaluation of seizures and epilepsy. There is limited information about its current use in the initial evaluation of pediatric epilepsy and about its yield during the initial diagnosis of epilepsy. We describe the patterns in the use and yield of diagnostic imaging in children with newly diagnosed epilepsy in a community-based study.

Methods.  Children were recruited when first diagnosed with epilepsy by participating physicians in Connecticut (1993-1997). Definitions for etiology and underlying epilepsy syndromes are as published by the International League Against Epilepsy.

Results.  Of 613 children, 488 (79.6%) had imaging: 388 (63.3%) magnetic resonance imaging, 197 (32.1%) computed tomography scans, and 97 (15.8%) both. Half of children with idiopathic generalized epilepsy had imaging studies compared with 70% to 100% of children with other forms of epilepsy, depending on the specific type. Etiologically relevant abnormalities were found in 62 (12.7% of those imaged). Fourteen of these children had otherwise completely normal presentations and histories. Their abnormalities included tuberous sclerosis (N = 4), tumors (N = 2), an arteriovenous malformation later diagnosed as a tumor, a cavernous angioma, cerebral malformations (N = 3), and other abnormalities (N = 3). Thirteen of the 14 had partial seizures and 12 had focal electroencephalographic (EEG) findings. Only 1 had neither.

Conclusions.  In children with newly diagnosed epilepsy, neuroimaging reveals a small but significant number of serious abnormalities not previously suspected. Most of these children have partial seizures or focal EEG abnormalities. Neuroimaging should be considered during the evaluation of children with newly diagnosed epilepsy, especially for those with neurologic deficits or partial seizures or focal EEG abnormalities that are not part of an idiopathic localization-related epilepsy syndrome.  Key words:  neuroimaging, epilepsy, children, evaluation, MRI.

Neuroimaging is an important tool in the evaluation and diagnosis of epilepsy.¹ In the United States, there is a high reliance on imaging in emergency department settings.² Expert opinion supports its use as part of the initial diagnosis and evaluation of epilepsy in children.^3-6 There is little information about the yield of diagnostic imaging in children with newly diagnosed epilepsy. The following describes patterns in the use of imaging and the results of a community-based study of newly diagnosed epilepsy in children in the 1990s.

	METHODS

Top Abstract Methods Results Discussion References

Children were prospectively recruited from 1993-1997 when they were first diagnosed with epilepsy (at least 2 unprovoked seizures occurring 1 day apart⁷ by participating physicians in Connecticut. The participating physicians included 16 of the 17 child neurologists in the state as well as several pediatricians and adult neurologists who, when canvassed, said that they would occasionally provide all neurologic care for a child with new onset epilepsy without referring to a child neurologist. All other physicians indicated that they always sought a pediatric neurology consult for a child with suspected epilepsy. In each practice, methods for active surveillance were put in place during the recruitment interval so that all potentially eligible candidates for the study would be identified. Eligible children were 1 month to 15 years at the time of the first unprovoked seizure. The investigators classified epilepsy syndromes and seizure types as well as etiology based on all available medical records and information obtained through interviewing the parents.^8,9 Definitions for etiology, epilepsy syndromes, and seizure types as published by the International League Against Epilepsy were used.^7,10,11 Idiopathic etiology is reserved for idiopathic syndromes. Cryptogenic refers to nonidiopathic syndromes in patients with no neurologic abnormalities or conditions to which the occurrence of epilepsy may be attributed. Remote symptomatic refers to epilepsy in the presence of a neurologic abnormality associated with an increased risk of epilepsy (eg, cerebral palsy). The neurologic examination was considered abnormal if a definite neurologic abnormality was present. Children with only mild clumsiness, mild hypotonia with normal strength, or other soft neurologic signs were classified as having neurologic soft signs.

This was a strictly observational study. Consequently, imaging studies were ordered for clinical purposes by treating physicians at facilities throughout the state. Parents provided written informed consent and children written or oral assent as appropriate. Further details of the identification and recruitment of this cohort have been described previously.⁸

Neuroimaging studies were considered if they clearly were ordered as part of the initial diagnostic workup for epilepsy or had been performed before the onset of epilepsy but then information from them was used in the evaluation of epilepsy. This is reasonable as there would have been no point in most of these instances to repeat an earlier study unless a new insult or a progressive process was suspected. A magnetic resonance imaging (MRI) performed as recommended follow-up to a computed tomography (CT) scan or electroencephalogram (EEG) obtained as part of the initial evaluation was also considered part of the initial evaluation even if there was a slight delay before it was performed. Studies performed only in response to persistent recurrent seizures early in the period after diagnosis were not considered part of the initial diagnostic evaluation.

Statistical analysis was conducted with t tests for continuously distributed variables (eg, age) and ² tests for discrete variables. Logistic regression was used for multivariable analysis.¹²

	RESULTS

Top Abstract Methods Results Discussion References

In all, 613 children were recruited. The median age at onset was 5.3 years, and 307 (50%) were boys. A total of 488 (79.6%) children had neuroimaging studies. CT scans were performed in 197 (32.1%) and MRIs in 388 (63.3%). Ninety-seven (15.8%) children had both.

Which Children Have Neuroimaging?

Age at first unprovoked seizure was not substantially different in children with and without imaging studies (5.7 vs 6.2 years; P = .30). Neuroimaging was performed in 109 of 185 (58.9%) children ultimately classified with idiopathic etiology, 277 of 317 (87.4%) children classified with cryptogenic etiology; and 102 of 111 (91.9%) of those classified with remote symptomatic etiology. The underlying epilepsy syndrome was strongly correlated with whether imaging was performed (Table 1). Only 49.2% of children with idiopathic generalized epilepsies (IGE) had neuroimaging compared with an average 87% (range: 78%-100%) of children with other forms of epilepsy (P < .001).

Within the IGE group (N = 126), 2 factors distinguished children with and without imaging studies. Twenty-six of 32 (81.3%) who had generalized tonic-clonic seizures (GTCS) compared with 36 of 94 (38.3%) of those without GTCS had imaging (P < .001). In addition, of the 94 children who did not have GTCS, 18 of 36 (50.0%) of children with focal epileptiform abnormalities on their EEGs (a finding that can be seen in IGE¹³), but only 18 of 58 (31.3%) of those without focal features had imaging studies (P = .07). Other clinical factors associated with the likelihood of having an imaging study are presented in Table 2.

Frequency of Significant Imaging Abnormalities

In 62 children (12.7% of those imaged), scans revealed abnormalities of presumed etiologic relevance (significant) to their epilepsy (Tables 1 and 2). This represents 14.2% of those without generalized idiopathic syndromes. Children in the cryptogenic/symptomatic generalized category were particularly likely to have imaging abnormalities (~30% overall). In only 5 instances, the classification of the syndrome was changed because of the imaging findings. In 1 case, the working diagnosis based on EEG and clinical presentation had been benign rolandic epilepsy. Imaging evidence of a cyst changed the primary epilepsy diagnosis to symptomatic localization-related epilepsy. This suggests that the syndrome can be identified almost entirely from the EEG, seizure history, and clinical findings.

Indicators of Significant Imaging Abnormalities: Overall Group

Several clinical factors were associated with whether an etiologically significant imaging abnormality was present. In the overall group who had imaging (N = 488; Table 2), children with abnormalities were slightly younger at the onset of epilepsy than were children without (4.9 years vs 5.9 years; P = .08). Imaging abnormalities were especially common in children with abnormal motor exams (57.6%), but not with neurologic soft signs (2.3%). Abnormalities were also more common in association with status epilepticus and partial seizures as well as with focal spikes and focal slowing on the EEG. Only 1 child with an abnormality had none of these features. In this case, the EEG was normal. At study entry there was evidence suggestive of a partial onset of seizures; however, ultimately the consensus classification of the seizure type was generalized.^8,9

On multivariable analysis with logistic regression, an abnormal motor examination was the strongest predictor of abnormal imaging (odds ratio [OR]: 18.9; 95% confidence interval (CI): 9.9, 36.3; P < .0001). After adjustment for the neurologic examination, focal slowing on the EEG was independently associated with the likelihood of having an abnormal imaging study (OR: 2.7; 95% CI: 1.2, 6.0; P = .02). Within this overall group, the effect of partial seizures was overwhelmed by the neurologic examination. In addition, large proportions of the children with cryptogenic/symptomatic generalized syndromes had imaging abnormalities and generalized seizures (eg, in association with infantile spasms). Consequently, this served to decrease the magnitude of association for seizure type.

Indicators of Imaging Abnormalities in Otherwise Normal Children

There were 400 children whose initial clinical presentation was otherwise entirely normal. Of these, 14 (3.5%) had etiologically significant imaging abnormalities (Table 2). The age at onset was slightly greater in children with imaging abnormalities than in those without (7.2 vs 6.1 years; P = .30). Thirteen of 14 (92.3%) with imaging abnormalities had partial seizures. Children with imaging abnormalities were more also likely to have focal spikes, slowing or both on the EEG. With only 1 exception, all children who otherwise appeared to be normal but who had significant imaging abnormalities had either partial seizures or focal EEG findings. There was no association between imaging abnormalities and status epilepticus in this group.

In a multiple logistic regression analysis, focal slowing (OR: 4.4; 95% CI: 1.4, 14.3; P = .01) and partial seizures (OR: 7.8; 95% CI: 1.0, 61.1; P = .05) were independent indicators of a significant imaging abnormalities. In the 31 children with both focal slowing and partial seizures, 5 (16.1%) had significant imaging abnormalities.

Specific Abnormalities

In 36 children with imaging abnormalities, imaging had been initiated before the onset of epilepsy (ie, the first unprovoked seizure) or for other indications. Of the 26 children with abnormalities who were imaged as part of their initial evaluation for unprovoked seizures, 14 had completely normal examinations and unrevealing histories. Of these 14, 4 had tuberous sclerosis, 2 had tumors (dysembryoplastic neuroepithelial tumor [DNET], and ganglioglioma), and 1 had what was initially diagnosed as an arteriovenous malformation (AVM) but was later found to be an astrocytoma. Other abnormalities included 3 brain malformations, 1 cavernous angioma, 1 venous angioma, 1 cystic lesion, and 1 instance of an apparent intrauterine insult.

Surgery

Two children (1 with a cavernous angioma, 1 with a ganglioglioma) had surgery within 1 to 2 months of diagnosis. Neither child had any seizures after initial diagnosis and drug treatment. The decision to operate was based on the lesion itself. At the time of initial diagnosis, it was recommended that the child with the DNET have an excisional biopsy or full resection. Because of the location of the lesion in the supplemental motor area, the child was closely monitored instead. After failing the first drug, he had a resection. Two other children had surgery. The child with the AVM failed 2 drugs and then underwent a surgical evaluation that indicated a tumor rather than an AVM. The pathology report identified the lesion as a grade II astrocytoma. In addition, 1 child diagnosed with presumed mesial temporal sclerosis had persistent auras and eventually underwent resective surgery. Analysis of the resected tissues revealed an oligodendroglioma.

CT Versus MRI

In 18 of the 62 children with etiologically relevant abnormalities, both a CT and MRI were performed. In 15 instances, the abnormality was identified on CT and confirmed on MRI. In 3 instances, the CT was normal and the MRI abnormal. The findings that were missed by CT were atrophy, forme fruste of schizencephaly, and an abnormal periventricular white matter signal in a child with cerebral palsy.

Incidental Abnormalities

In addition to the above etiologically significant abnormalities, incidental abnormalities that were not considered relevant to underlying etiology (4 pineal cysts, 2 choroid fissure cysts, 5 mild Chiari I malformations, 3 arachnoid cysts, and 1 instance of multiple tiny clustered foci in the thalamus) or insufficient basis for classifying etiology as remote symptomatic (hippocampal atrophy, N = 3) occurred in 18 children (Table 3).

Subsequent Imaging After Initial Evaluation Period

These data are from an ongoing study, and information continues to accumulate. The following summarizes the findings of subsequent neuroimaging during the first 2 years after study entry.

No Imaging at Baseline

To date, of the 125 who had no imaging at baseline, 11 have had subsequent studies (10 MRI and 1 CT). Eight of the 11 studies were completely normal. One demonstrated an old infarct in a child with an abnormal neurologic examination. Another demonstrated some mild hippocampal asymmetry. The third study was performed specifically because of headaches, not seizures, and revealed a Chiari I malformation, not considered relevant to the seizure disorder.

Equivocal Findings at Baseline

Of the 65 children whose baseline neuroimaging evaluations were interpreted as equivocal (considered normal in the above presentation), 17 have had subsequent MRIs of which 6 were normal and 11 were not. Of the abnormal scans, 1 revealed a neuromigrational defect (not fully appreciated on baseline CT); 1 revealed an area of possible encephalomalacia with interval volume loss; 1 revealed a cystic lesion; 3 revealed delayed myelination; and 2 revealed a hyperintensity unchanged from the baseline study and of unclear significance. One child's follow-up scans revealed the development of hippocampal atrophy. In 2 others, subsequent studies indicated changes consistent with a progressive encephalopathy.

CT Only at Baseline

Of the children who had only CT scans at baseline, 14 have had subsequent studies during the first 2 years of follow-up. Seven of these (5 MRI and 2 CT) were normal. Three (2 MRI, 1 CT) confirmed abnormalities already revealed by initial CT. Of the remaining 4 MRIs, 1 demonstrated a hippocampal asymmetry; 1 revealed a venous angioma not considered relevant to that particular child's epilepsy; 1 indicated a neuromigrational defect, and the last revealed progressive volume loss associated with a neurodegenerative condition. The last 2 are also mentioned above under follow-up to equivocal findings.

	DISCUSSION

Top Abstract Methods Results Discussion References

To determine the utility of imaging and provide information relevant to developing guidelines for its use in children with newly diagnosed epilepsy, it is necessary to study a broad unselected spectrum of patients. Because of the high reliance in Connecticut on neurodiagnostic imaging in the evaluation of newly recognized pediatric epilepsy, we were largely able to do this. We were also able to describe some of factors that influenced whether or not imaging was used. Specifically, it was used much less for IGE. When it was used, it tended to be in children who had GTCS or evidence of focality on the EEG the occurrence of either of which can cast some uncertainty on the diagnosis of generalized idiopathic epilepsy.

The yield of neuroimaging in children in this study is similar to earlier reports that focused on CT scans but which were not always able to describe the selection factors operating on the case series^14,15 and to a preliminary report of a hospital-based series in which MRI was used.¹⁶ The yield of abnormalities by epilepsy syndrome is also comparable to another study from Australia that found no imaging abnormalities in children with EEG-confirmed idiopathic epilepsies.¹⁷ The findings for yield are also comparable to those of children with a first unprovoked seizure¹⁸ even though the types of epilepsy that present with a first seizure tend to differ from those that present as epilepsy.¹⁹ Our figure of 12.4% for etiologically relevant abnormalities is much lower, however, than in studies of children with long-standing epilepsy^20-22 and adults.^23-25 This is not surprising, as children with brain abnormalities are most likely to be disproportionately overrepresented in series with long-standing, refractory epilepsy. Adult epilepsy is also more frequently secondary to underlying abnormalities.

Neuroimaging has the highest yield in children with abnormal neurologic exams. Partial seizures and focal EEG abnormalities further identify children with imaging abnormalities. Similar findings have been reported in other series of patients.^{15,17,18,24,26,27} Only 14 (3.5%) otherwise normal children had significant imaging abnormalities, and approximately 1% had serious lesions that altered immediate management. In apparently normal children, partial seizures and focal EEG findings have a high but not 100% sensitivity for neuroimaging abnormalities.

We caution that our study does not address criteria for performing emergency imaging in children presenting to emergency departments with new onset seizures.²⁸ In this circumstance, the diagnostic issues include identification of acute provocations that require immediate intervention (eg, cerebral hemorrhage). Patients whose only seizures were caused by acute provocations were not eligible for our study.⁷

A limitation of our study is the lack of uniform imaging protocols. Not all children were imaged, and the techniques used may have differed between facilities. However, the group represents a newly diagnosed community-based series with recruitment from all but 1 child neurologist in the state, and we were able to describe factors associated with having imaging. Ideally, for research purposes, all children would have been imaged. Potentially, in those not imaged, there may be structural abnormalities present, and our designation of syndrome and etiology is incorrect. On the other had, those children least likely to be imaged are those who meet the clinical and EEG criteria for IGE, a group unlikely to be mistakenly diagnosed. Subsequent imaging during the first 2 years of follow-up in this cohort has revealed relatively little that was missed at initial diagnosis and nothing that would have altered immediate treatment.

Given the patterns of care in other countries, it is doubtful that imaging is used as frequently in the evaluation of newly diagnosed pediatric epilepsy as in Connecticut.^29-32 As 90% of those without clearly idiopathic syndromes, based on history, examination, and EEG were imaged, this is likely to be one of the most complete surveys of neuroimaging findings in newly diagnosed pediatric epilepsy.

Whether imaging should always be performed is controversial. New onset seizures sometimes signal the presence of a serious brain abnormality, and early identification and treatment may improve long-term outcome. On the other hand, the cost of performing MRIs in all children with new onset epilepsy would be substantial. Some of the cost of evaluating pediatric epilepsy could be reduced if the use of emergency CT scans were curtailed in favor of a more definitive MRI.²⁰ This, of course, first requires a separate evaluation of CT use in the emergency department setting to identify acute conditions that require immediate treatment. In addition, if a thorough history were taken and neurologic examination and EEG were performed first, this could allow the screening out of individuals with clearly idiopathic epilepsies and result in some reduction of the numbers of children referred for any imaging. Other factors of course may influence a clinician's decision to recommend imaging. These include the need for sedation depending on the child's age.

The outcome of nonidiopathic epilepsy, most of which is localization-related⁸ depends, in part, on etiology.³³ In this group, there is a substantial yield of abnormalities that affect prognosis and influence long-term treatment and management strategies. In addition, there are a small but important number of serious abnormalities (affecting ~1%) that influence immediate treatment. For these reasons we conclude that the findings of our study support published recommendations that neuroimaging be considered when evaluating children with new onset seizures unless an idiopathic syndrome can be clearly demonstrated. This is especially true in the face of neurologic abnormalities or partial seizures or focal EEG findings that are not part of an idiopathic localization-related epilepsy syndrome.^3-6

	ACKNOWLEDGMENTS

This study was funded by a Grant NINDS RO1 NS31146 from the National Institutes of Health.

We are especially grateful to the parents and children who patiently and selflessly have participated in this study. We also extend out thanks to the other physicians in Connecticut who have referred their patients to this study, Drs Robert Cerciello, Francis Dimario, Barry Russman, Michelle Kleiman, Carol Leicher, Edward Zalneraitis, Philip Brunquel, Laura Ment, Edward Novotny, Bennet Shaywitz, S. Nallainathan, Alok Bargava, Martin Kreminitzer, Barbara Coughlin, Harriet Fellows, Jack Finkelstein, Daniel Moalli, Louise Resor, Brenda Balch, Patricia Braun, Owen Erlich, Bernard Giserman, John Monroe, Lawrence Rifkin, Lourdes Rosales, and Murray Engel. We also thank Drs Edward Novotny and Francis DiMario for reinterpreting selected EEGs for the study. Dr Eugene Shapiro has kindly facilitated many administrative issues for us. We also thank the research associates, Susan Smith-Rapaport, Barbara Beckerman, Heather LaCoste, Lynnette Bates, Joann Gehrels, and Kris Engel for their dedicated work on this project and Wuthikrai Uayingsak for his exceptional programming expertise.

	FOOTNOTES

Received for publication Aug 9, 1999; accepted Dec 23, 1999.

Reprint requests to (A.T.B.) Department of Biological Sciences, Northern Illinois University, DeKalb, IL 60115. E-mail: atberg{at}niu.edu

	ABBREVIATIONS

MRI, magnetic resonance imaging; CT, computed tomography (scan); EEG, electroencephalogram; IGE, idiopathic generalized epilepsies; GTCS, generalized tonic-clonic seizures; OR, odds ratio; CI, 95% confidence interval; DNET, dysembryoplastic neuroepithelial tumor; AVM, arteriovenous malformation.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Cook M, Kilpatrick C Imaging in epilepsy. Curr Opin Neurol. 1994; 7:123-130 [Medline]
2. Greenberg MK, Barsan WG, Starkman S Neuroimaging in the emergency patient presenting with seizure. Neurology. 1996; 47:26-32 [Free Full Text]
3. Gilliam F, Wyllie E Diagnostic testing of seizure disorders. Neurol Clin. 1996; 14:61-84 [CrossRef][Medline]
4. Holmes GL How to evaluate the patient after a first seizure. Postgrad Med. 1988; 83:199-209
5. Kuzniecky RI Neuroimaging in pediatric epilepsy. Epilepsia. 1996; 37:S10-S21
6. Nordli DR, Pedley TA. Evaluation of children with seizures. In: Shinnar S, Amir N, Branski D, eds. Childhood Seizures. Basel, Switzerland: S Karger; 1995:67-77
7. Commission on Epidemiology and Prognosis, International League Against Epilepsy Guidelines for epidemiologic studies on epilepsy. Epilepsia. 1993; 34:592-596 [CrossRef][Medline]
8. Berg AT, Shinnar S, Levy SR, Testa FM Newly-diagnosed epilepsy in children: presentation at diagnosis. Epilepsia. 1999; 40:445-452 [CrossRef][Medline]
9. Berg AT, Levy SR, Testa FM, Shinnar S Classification of childhood epilepsy syndromes in newly diagnosed epilepsy: interrater agreement and reasons for disagreement. Epilepsia. 1999; 40:439-444 [CrossRef][Medline]
10. Commission on Classification and Terminology of the International League Against Epilepsy Proposal for revised clinical and electrographic classification of epileptic seizures. Epilepsia. 1981; 22:489-501 [Medline]
11. Commission on Classification and Terminology of the International League Against Epilepsy Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia. 1989; 30:389-399 [Medline]
12. SAS Institute, Inc. SAS/STAT User's Guide, Version 6. 4th ed. Cary, NC: SAS Institute, Inc; 1990
13. Aicardi J. Epilepsy in Children. New York, NY: Raven Press; 1986
14. Gilsanz V, Strand R, Barnes P, Nealis J Results of presumed cryptogenic epilepsy in childhood by CT scanning. Ann Radiol. 1979; 22:184-187
15. McAbee GN, Barasch ES, Kurfist LA Results of computed tomography in neurologically normal children after initial onset of seizures. Pediatr Neurol. 1989; 5:102-106 [CrossRef][Medline]
16. DeGrauw TJ, Franz DN Brain mass lesions detected by MRI in children with new onset seizures. Epilepsia. 1997; 38:124 [CrossRef][Medline]
17. King MA, Newton MR, Graeme GD, Epileptology of the first seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive cases. Lancet. 1998; 352:1007-1011 [CrossRef][Medline]
18. O'Dell C, Shinnar S, Mitnick R, Berg AT, Moshe SL Neuroimaging abnormalities in children with a first afebrile seizure. Epilepsia. 1997; 38:184
19. Shinnar S, O'Dell C, Berg AT Epilepsy syndromes in children identified at the time of their first unprovoked seizure. Epilepsia. 1998; 39:148
20. Resta M, Palma M, Dicuonzo F, Imaging studies in partial epilepsy in children and adolescents. Epilepsia. 1994; 35:1187-1193 [CrossRef][Medline]
21. Yang PJ, Berger PE, Cohen ME, Duffner PK Computed tomography and childhood seizures disorders. Neurology. 1979; 29:1084-1088 [Abstract/Free Full Text]
22. Bachman DS, Hodges F, Freeman JM Computerized axial tomography in chronic seizure disorders of childhood. Pediatrics. 1976; 58:828-832 [Abstract]
23. Schoenenberger RA, Heim SM Indication for computed tomography of the brain in patients with first uncomplicated generalised seizure. Br Med J. 1994; 309:986-989 [Abstract/Free Full Text]
24. Ramirez-Lassepas M, Cipolle RJ, Morillo LR, Gumnit RJ Value of computed tomographic scan in the evaluation of adult patients after their first seizures. Ann Neurol. 1984; 15:526-543
25. Daras M, Tuchman AJ, Zisfein J Use of CT in epilepsy: does it matter how long you wait? Eur Neurol. 1991; 31:428-429 [Medline]
26. Harwood-Nash DC Computed tomography and seizures in children. J Neuroradiol. 1983; 10:130-136 [Medline]
27. Galli V, Benatti A, Venuta A, Rozzi N, Crisi G The role of computerized tomography in the epilepsies of childhood. Med Surg Pediatr. 1985; 7:1-8
28. Warden CR, Brownstein DR, Del Beccaro MA Predictors of abnormal findings of computed tomography of the head in pediatric patients presenting with seizures. Ann Emerg Med. 1997; 29:518-523 [CrossRef][Medline]
29. Henriksen O Perspectives of epilepsy care in Norway. Neurology. 1997; 48:S20-S24 [Medline]
30. Dulac O, Jallon P Patterns of care for patients with epilepsy in France. Neurology. 1997; 48:S30-S32 [Abstract]
31. Chadwick D, Reynolds EH Services for epilepsy in the United Kingdom. Neurology. 1997; 48:S3-S7 [Medline]
32. Schmidt D Epilepsy care in Germany: a clinical perspective. Neurology. 1997; 48:S25-S29 [Medline]
33. Sillanpaa M, Jalava M, Kaleva O, Shinnar S Long-term prognosis of seizures with onset in childhood. N Engl J Med. 1998; 338:1715-1722 [Abstract/Free Full Text]