pISSN 2234-8999 / eISSN 2288-2243

CLINICAL ARTICLE
Korean J Neurotrauma 2016;12(2):112-117 https://doi.org/10.13004/kjnt.2016.12.2.112

Effectiveness of Cortical Atrophy Scale and Indirect Indices

of Brain Atrophy to Predict Chronic Subdural Hematoma
in Older Patients
Eun-Oh Jeong, MD, Seung-Won Choi, MD, Jeong-Wook Lim, MD, Hyon-Jo Kwon, MD,
Seon-Hwan Kim, MD, Hyeon-Song Koh, MD, Jin-Young Youm, MD, and Shi-Hun Song, MD
Department of Neurosurgery, School of Medicine, Chungnam National University Hospital, Daejeon, Korea

Objective: To determine whether baseline cerebral atrophy can predict the rate of future chronic subdural hematoma
(CSDH) after head trauma and compare indirect markers of brain atrophy with volumetric analysis of computed tomogra-
phy (CT).
Methods: Single institution case-control study involving 1,476 patients who visited our hospital after head trauma from
January 2009 to December 2014. Forty-one patients with delayed CSDH were identified and age, gender matched 41 pa-
tients were selected as control group. Both volumetric analyze on CT and Evans index, frontal horn index, bicaudate ratio,
sylvian fissure ratio and cortical atrophy scale of 82 patients were estimated by different raters and relationship of those
factors with CSDH was analyzed.
Results: Every indirect indices except cortical atrophy scale were not enough to give a good estimate of CSDH. Brain at-
rophy and cortical atrophy scale were predisposing factors of CSDH on multivariate analysis with statistical significance.
Conclusion: Brain atrophy was a potential prognostic factor of CSDH after trauma. In practice, patients with a value of
cortical atrophy scale over moderate grade needed more attention for CSDH.
(Korean J Neurotrauma 2016;12(2):112-117)

KEY WORDS: Analysis ㆍAtrophy ㆍBrain ㆍEvaluation studies as topic ㆍHematoma, subdural, chronic.

Introduction the mechanisms of development, evolution, recurrence and

risk factors of CSDH have been identified but needed to be
Chronic subdural hematoma (CSDH) is one of the most confirmed. Brain atrophy is one of the high risks of CSDH
common types of intracranial (IC) hemorrhage that occurs especially in old age. Volume quantification of brain par-
at a rate of 1 to 2 per 100,000 per year.1,5,16) With advances enchyme may provide better description of brain atrophy,
in computed tomography (CT), brain scans, magnetic reso- as well as prediction of CSDH after head trauma, but it can
nance imaging (MRI), and molecular biologic technologies, be difficult and time consuming for all patients in practice.
There are many kinds of traditional indirect indices of
Received: August 19, 2016 / Revised: October 4, 2016 brain atrophy but correlation between indirect indices and
Accepted: October 5, 2016 CSDH is little known.5,12,13) In this study, we evaluated brain
Address for correspondence: Seung-Won Choi
Department of Neurosurgery, School of Medicine, Chungnam atrophy of patients with head trauma by indirect indices
National University Hospital, 282 Munhwa-ro, Jung-gu, Daejeon (qualitative value) and its’ influence on CSDH. We tried to
35015, Korea
Tel: +82-42-280-7361, Fax: +82-42-280-7364 find reference value of indirect indices for predicting CSDH
E-mail: [email protected] and to correlate indirect indices and volumetric analysis of
cc This is an Open Access article distributed under the terms of Cre-
brain CT (quantitative value).
ative Attributions Non-Commercial License (http://creativecommons.
org/licenses/by-nc/3.0/) which permits unrestricted noncommercial use,
distribution, and reproduction in any medium, provided the original work
is properly cited.

112 Copyright © 2016 Korean Neurotraumatology Society

 Eun-Oh Jeong, et al.

Materials and Methods rophy index (%)=(1-CSF volume/IC volume)×100] (Figure 1).

Patients Indices of brain atrophy (qualitative value)

A retrospective study was carried out on head trauma Similar to volumetric analysis, 3 raters measured the fol-
patients admitted in our hospital from January 2009 to De- lowing 5 traditional indirect indices of brain atrophy.
cember 2014. Initially, 1,476 head trauma patients were 1) Evans index (EI)-Typically used as a measure of ven-
found and 530 patients were over 65 years old. 304 patients tricle volume. EI is measured as the ratio of the maximal
of 530 patients were finally remained after following exclu- width of the frontal horns to internal width of cranium.3)
sion criteria; Both diameters were measured in the same CT slice.
1) Patient with initially diagnosed IC hematoma who re- 2) Frontal horn index (FHI)-FHI was measured as the
quired surgery at admission. ratio between the maximal width of both frontal horns and
2) Patients who underwent surgery within 7 days due to the width of the whole brain at the same level.15)
hematoma aggravation after head trauma. 3) Bicaudate ratio (BCR)-Typically used as a measure of
3) Patients with initial Glasgow Coma Scale under 12. internal cerebral atrophy. The axial slice on which the cau-
4) Patients who had IC space-occupying lesion or brain date nuclei produced the greatest amount of indentation
edema. on the lateral ventricles was identified, and the distance be-
5) Patients who had delayed CSDH within 3 weeks. tween the 2 caudate apices was measured. This value was
Among 304 patients who were not operated at acute or divided by the maximum width of the skull at the same lev-
subacute phase, 41 patients had CSDH over 3 weeks after el as the caudate measurement.2,14)
initial head trauma and classified as Group A. Gender and 4) Sylvian fissure ratio (SFR)-Typically used as a measure
age matched (±5 years) 41 patients of 263 patients were of external cerebral atrophy. The average of the maximum
randomly selected as a control group and classified as width of the 2 sylvian fissures on the section showing them
Group B. All patients’ medical records were reviewed in- at their widest was divided by the trans-pineal coronal inner
cluding history of diabetes mellitus (DM), hypertension table diameter.14)
(HTN), cerebral infarction, cardiac disease, hematologic 5) Cortical atrophy scale-It was devised by Scheltens and
disease and epilepsy. colleagues.11) Scores 0-3 represent absent, mild, moderate
and severe cortical atrophy, respectively. Mild changes are
Volumetric analysis of CT (quantitative value) said to be present when there is sulci opening peripherally.
Volume measurement of Gamma-Plan program (Elekta Moderate changes are seen when there is widening along
instrument AB, Stockholm, Sweden) by sequential volume the length of the sulcus. Severe atrophy is present when
mapping was used to access quantitative volume of brain there is gyrus thinning. In this article, we assumed that pa-
except cerebellum. Each slice was blindly analyzed by 2 tients had cortical atrophy when cortical atrophy scale was
neurosurgeons and 1 medical physicist and final evaluation over moderate grade (Figure 2).
was made by consensus between 3 raters. Brain atrophy in-
dex was defined as the ratio of cerebrospinal fluid (CSF) Statistical analysis
volume to IC volume as expressed in percentage.4,19) [brain at- Statistical analysis was carried out with the Windows ver-

FIGURE 1. Volume measurement of

Gamma-Plan program (Elekta instru-
ment AB, Stockholm, Sweden) by se-
quential volume mapping. (A) Black line
contain Intracranial volume. (B) Black A B
lilne contain cerebrospinal fluid volume.

http://www.kjnt.org 113
Cortical Atrophy Scale and Indirect Indices to Predict CSDH

A B C D
FIGURE 2. Brain computed tomography examples of cortical atrophy scale. (A) Absent atrophy. (B) Mild atrophy. (C) Moderate at-
rophy. (D) Severe atrophy.

TABLE 1. Univariate analysis of factors on chronic subdural hematoma after trauma

Factors Group A (n=41) Group B (n=41) Univariate analysis, p-value
Age 73.7 (±6.5) 78.2 (±7.1) 0.059
Gender 13/38 15/35 0.832
Diabetes mellitus 12 06 0.130
Hypertension 28 21 0.195
Cerebral Infarction 13 06 0.083
Cardiac disease 11 08 0.474
Hematologic disease 01 02 0.546
Epilepsy 01 00 0.816
Antiplatelet agent 19 17 0.522
Anticoagulant agent 04 02 0.414
Smoking 08 14 0.154
Alcohol 06 08 0.692

sion of SPSS Version 20.0 (SPSS Inc., Chicago, IL, USA). Correlation between indirect indices and quantitative
Student t-test and chi-square were utilized for difference volume
between groups; in addition, linear regression was used to BCR and EI had significant and moderately strong cor-
quantify the relationship between brain volume and indi- relation with ventricle volume (r=0.679, r=0.44, p=0.000,
ces of brain atrophy. And linear regression with 95% confi- respectively). BCR (r=0.447, p=0.000) was the most signifi-
dence interval was used to identify the correlation between cant influencing factor for brain atrophy, followed by EI
CSDH and indices of brain atrophy. A p-value of less than (r=0.247, p=0.17), and SFR (r=0.221, p=0.033). BCR also
was considered as statistically significance. had correlation with volume of cortical subarachnoid space,
but correlation coefficient was low (r=0.247, p=0.02). On
Results the other hand, there was no relationship between indirect
indices and brain hemisphere volume. All results were
Patient characteristics shown in Table 2.
The mean age was 73.7 years Group A, and 78.2 years in
Group B (p=0.059). Among 82 patients, 36 patients took an- Volumetric analysis (quantitative value)
tiplatelet agents; 19 patients of Group A and 17 patients of In quantitative study, brain volume was 1,158.1 mL in
Group B and there was no statistically significant (p=0.522). Group A and 1,130.6 mL in Group B (p=0.45). Ventricle vol-
History of cerebral infarction showed higher percentage in ume was 50.8 mL in Group A and 42.5 mL in Group B and
Group A than Group B, with borderline significance (p= volume of cortical subarachnoid space was 89.2 mL in
0.083). Gender, DM, HTN, smoking, anticoagulation agent Group A and 85.2 mL in Group B with no statistically sig-
and history of heart disease, epilepsy, hematologic disease nificance, respectively (p=0.19, p=0.79). There was a high-
had no statistical significance between 2 groups. We sum- er degree of brain atrophy index in Group A (9.08%) than
marized all results in Table 1. Group B (6.38%), with statistical significance (p=0.003). All

114 Korean J Neurotrauma 2016;12(2):112-117

 Eun-Oh Jeong, et al.

results are shown in Table 3. tween Group A and B (p=0.014). In brain cortical atrophy
(2 or 3 of cortical atrophy scale), there was 17 patients in
Indirect indices analysis (qualitative value) Group A and 6 patients in Group B with statistical signifi-
EI was 0.24 of Group A and 0.25 of Group B and there cance (p=0.024).
was no statistical significance (p=0.348). FHI was 0.29 of
Group A and 0.3 of Group B without statistical significance. TABLE 4. Multivariate analysis of factors on chronic subdural
In BCR and SFR, there was higher tendency in Group A hematoma after trauma

than Group B but, there was no statistical significance. On Factors p-value Exp (ß) 95% CI

the other hand, cortical atrophy scale and brain cortical at- Brain atrophy index* 0.029 3.1 1.092-11.853
Brain cortical atrophy† 0.042 2.02 0.721-4.8930
rophy had statistical significance (p=0.014, p=0.024, re-
spectively). All results are shown in Table 3. *brain atrophy index: 1-brain volume/intracranial volume ×100%,
†brain cortical atrophy: value of brain cortical atrophy scale

Cortical atrophy scale had statistically significant be- over moderate grade. CI: confidence interval

TABLE 2. Pearson correlation between indirect indices and quantitative volume

Brain Ventricle Evans Frontal Bicaudate Sylvian
Factors BAI*
volume volume index horn index ratio fissure ratio
BAI Pearson corrlation (r) 1
Sig. (2-tailed)
Brain Pearson corrlation (r) -0.225 1
volume Sig. (2-tailed) p=0.03
Ventricle Pearson corrlation (r) 0.679 -0.124 1
volume Sig. (2-tailed) p=0.000 p=0.236
Evans index Pearson corrlation (r) 0.247 -0.025 0.438 1
Sig. (2-tailed) p=0.17 p=0.810 p=0.000
Frontal horn Pearson corrlation (r) -0.03 -0.039 0.126 0.211 1
index Sig. (2-tailed) p=0.773 p=0.710 p=0.227 p=0.034
Bicaudate Pearson corrlation (r) 0.447 -0.109 0.649 0.557 0.139 1
ratio Sig. (2-tailed) p=0.000 p=0.299 p=0.000 p=0.000 p=0.167
Sylvian fissure Pearson corrlation (r) 0.221 -0.119 0.190 0.172 -0.011 0.126 1
ratio Sig. (2-tailed) p=0.033 p=0.255 p=0.068 p=0.086 p=0.915 p=0.209
*brain atrophy index: 1-brain volume/intracranial volume ×100%. BAI: brain atrophy index

TABLE 3. Indirect indices and brain volume between chronic subdural hematoma and control groups
Factors Group A (n=41) Group B (n=41) p-value
Quantitative value
Brain (mL) 1158.1 1130.6 0.45
Ventricle (mL) 0050.8 0042.5 0.19
Cortical subarachnoid space (mL) 0089.2 0085.2 0.79
Brain atrophy index* (%) 0009.08 0006.38 0.003
Qualitative value
Evans index 0000.24 0000.25 0.348
Frontal horn index 0000.29 0000.30 0.357
Bicaudate ratio 0000.15 0000.14 0.417
Sylvian fissure ratio 0000.067 0000.056 0.181
Brain cortical atrophy scale 0.014
Normal 0007 0018
Mild 0017 0017
Moderate 0011 0006
Severe 0006 0000
Brain cortical atrophy † 0017 0006 0.024
*brain atrophy index: 1-brain volume/intracranial volume ×100%, †brain cortical atrophy: value of brain cortical atrophy scale
over moderate grade

http://www.kjnt.org 115
Cortical Atrophy Scale and Indirect Indices to Predict CSDH

Predisposing factors of CSDH Qualitative and quantitative volumetric analysis

The multivariate analysis of the potential predisposing There are several previous studies to estimate part or to-
factors for the CSDH was done and there was statistically tal volume of a structure or region of interest. Qualitative
significant in brain atrophy index (p=0.029) and cortical at- indices are useful in diagnosis and estimation of disease,
rophy scale (p=0.042). Results were shown in Table 4. but it is important to select an appropriate index depending
on disease and purpose.
Discussion In order to estimate correlation between brain atrophy and
indirect indices of brain atrophy, multiple linear regression
Patient characteristics analysis was conducted. BCR and SFR had positive corre-
In initial study design, we tried to reduce the effect of oth- lation with brain atrophy index with statistically signifi-
er factors on chronic subdural hemorrhage except for indi- cant but other indices had no correlation. However, both
rect indices. So, age and gender matched 41 patients were indices had low correlation coefficient with brain atrophy
randomly selected as the control group. In chi-square anal- index, so it seemed to be hard to give a good estimate of
ysis, there was no statistically significant factors between brain atrophy.
2 groups, and this can be considered that performance of EI is usually used to estimate ventricle volume especial-
random sampling was appropriate. ly for hydrocephalus.3) In 2005, the idiopathic normal pres-
sure hydrocephalus (INPH) guidelines9) recommended that
Brain atrophy a value of EI above 0.3 on CT or MRI was defined as ven-
Recently, increasing of older peoples and patients had an- tricular dilation. In our study, EI was even slightly higher in
tiplatelet or anticoagulant treatment, CSDH patients have 0.25 of control group than 0.24 of CSDH group. EI had posi-
been increased. There are many reports that attempted to tive correlation with quantitative ventricle volume but not
confirm predisposing factors, because CSDH patients are in brain atrophy index.
diagnosed after a few weeks from trauma. Among many Cortical atrophy scale is usually used in neurology to study
predisposing factors, brain atrophy has been reported as re- degenerative neuronal disease.6) In initial analysis, there
liable factor.5,12,19) was no patient in control group who had severe grade of cor-
Direct or indirect head trauma can cause injury of the tical atrophy scale. So it was hard to see there was mean-
bridging vein or Mittenzweig’s vessels, resulting in hema- ingful result of cortical atrophy scale between 2 groups. In
toma in the subdural space. Subsequently, acute hemato- this regard, we divided patients based on value of cortical
ma can undergo absorption through hemolysis or gradual atrophy scale over 2 or not and there was statistical signifi-
enlargement to cause symptoms. In this process, brain atro- cance between 2 groups. Möller et al.7) reported that corti-
phy allows ease of separation of the dura-arachnoid inter- cal atrophy scale was quantitatively validated and reliably
space on head trauma, with resultant fibrin deposits and ac- reflected brain atrophy compared with volumetric analy-
cumulation of serum and exudates that induce proliferation sis on MRI. In multiple regression analysis of 5 indirect in-
of granulation tissue on the inner dural surface. This pro- dices and brain atrophy index, cortical atrophy and brain
liferation of dura results in production of neo-membrane and atrophy index had statistical significant. Cortical atrophy
subsequent growth of new vessels directly within the sub- scale has shown to be a quick and reproducible method to
dural space. The repeated micro-bleeding from these ves- assess the degree of atrophy in practice.
sels cause CSDH.8,10,17,18) There were some limitations in our study due to its retro-
Yang et al.19) analyzed degree of brain atrophy in 19 pa- spective nature. In radiologic features, MRI can be more
tients of CSDH and 76 age-matched patients as control group appropriate for quantitative as well as qualitative evalua-
and reported that there was higher degree of brain atrophy tion, but brain CT is usually taken at the time of traumatic
in CSDH group. In their study, mean brain atrophy index brain injury. Intra-rater reliability for cortical atrophy scale
was 14.3% and even stronger predictor of CSDH especial- is high; however, there is a need to confirm inter-rater reli-
ly for younger patients under 65 years old. In our study an- ability.
alyzed older patients over 65 years old, mean brain atrophy
index was 9.08% of CSDH patients with statistical signifi- Conclusion
cant and also predisposing factors for the CSDH.
Brain atrophy is a potential prognostic factor of CSDH
after trauma. In practice, indirect indices are used to esti-

116 Korean J Neurotrauma 2016;12(2):112-117

 Eun-Oh Jeong, et al.

mate brain atrophy and cortical atrophy scale had better pre- ment of idiopathic normal-pressure hydrocephalus. Neurosurgery
57:S17-S28, 2005
diction of CSDH in qualitative and quantitative volumetric
10) Sajanti J, Majamaa K. High concentrations of procollagen pro-
analysis. Careful image interpretation is needed for early peptides in chronic subdural haematoma and effusion. J Neurol
diagnosis of CSDH. Neurosurg Psychiatry 74:522-524, 2003
11) Scheltens P, Pasquier F, Weerts JG, Barkhof F, Leys D. Qualitative
■ The authors have no financial conflicts of interest. assessment of cerebral atrophy on MRI: inter- and intra-observer
reproducibility in dementia and normal aging. Eur Neurol 37:95-
REFERENCES 99, 1997
12) Son S, Yoo CJ, Lee SG, Kim EY, Park CW, Kim WK. Natural
1) Ahmed E, Aurangzeb A, Khan SA, Maqbool S, Ali A, Zadran KK, course of initially non-operated cases of acute subdural hemato-
et al. Frequency of conservatively managed traumatic acute sub- ma: the risk factors of hematoma progression. J Korean Neuro-
dural haematoma changing into chronic subdural haematoma. J surg Soc 54:211-219, 2013
Ayub Med Coll Abbottabad 24:71-74, 2012 13) Tugcu B, Tanriverdi O, Baydin S, Hergunsel B, Günaldı Ö, Ofluo-
2) Bermel RA, Bakshi R, Tjoa C, Puli SR, Jacobs L. Bicaudate ratio glu E, et al. Can recurrence of chronic subdural hematoma be pre-
as a magnetic resonance imaging marker of brain atrophy in mul- dicted? A retrospective analysis of 292 cases. J Neurol Surg A
tiple sclerosis. Arch Neurol 59:275-280, 2002 Cent Eur Neurosurg 75:37-41, 2014
3) Evans W. An encephalographic ratio for estimating ventricular en- 14) van Zagten M, Kessels F, Boiten J, Lodder J. Interobserver agree-
largement and cerebral atrophy. Arch Neurol Psychiatry 47:931- ment in the assessment of cerebral atrophy on CT using bicaudate
937, 1942 and sylvian-fissure ratios. Neuroradiology 41:261-264, 1999
4) Goldszal AF, Davatzikos C, Pham DL, Yan MX, Bryan RN, Resn- 15) Walchenbach R, Geiger E, Thomeer RT, Vanneste JA. The value
ick SM. An image-processing system for qualitative and quanti- of temporary external lumbar CSF drainage in predicting the out-
tative volumetric analysis of brain images. J Comput Assist To- come of shunting on normal pressure hydrocephalus. J Neurol
mogr 22:827-837, 1998 Neurosurg Psychiatry 72:503-506, 2002
5) Jack A, O’Kelly C, McDougall C, Findlay JM. Predicting recur- 16) Winn HR. Youmans neurological surgery, ed 5. Philadelphia, PA:
rence after chronic subdural haematoma drainage. Can J Neurol W. B. Saunders, 2004
Sci 42:34-39, 2015 17) Yamamoto H, Hirashima Y, Hamada H, Hayashi N, Origasa H,
6) Koedam EL, Lehmann M, van der Flier WM, Scheltens P, Pijnen- Endo S. Independent predictors of recurrence of chronic subdural
burg YA, Fox N, et al. Visual assessment of posterior atrophy de- hematoma: results of multivariate analysis performed using a lo-
velopment of a MRI rating scale. Eur Radiol 21:2618-2625, 2011 gistic regression model. J Neurosurg 98:1217-1221, 2003
7) Möller C, van der Flier WM, Versteeg A, Benedictus MR, Wattjes 18) Yamazaki Y, Tachibana S, Kitahara Y, Ohwada T. Promotive fac-
MP, Koedam EL, et al. Quantitative regional validation of the vi- tors of chronic subdural hematoma in relation to age. No Shinkei
sual rating scale for posterior cortical atrophy. Eur Radiol 24:397- Geka 24:47-51, 1996
404, 2014 19) Yang AI, Balser DS, Mikheev A, Offen S, Huang JH, Babb J, et
8) Markwalder TM. Chronic subdural hematomas: a review. J Neu- al. Cerebral atrophy is associated with development of chronic sub-
rosurg 54:637-645, 1981 dural haematoma. Brain Inj 26:1731-1736, 2012
9) Marmarou A, Bergsneider M, Klinge P, Relkin N, Black PM. The
value of supplemental prognostic tests for the preoperative assess-

http://www.kjnt.org 117