Extent and Distribution ofExtent and Distribution ofWhite Matter Hyperintensities in White Matter Hyperintensities in

Normal Aging, Mild Cognitive Normal Aging, Mild Cognitive Impairment and AlzheimerImpairment and Alzheimer’’s Disease s Disease

Mitsuhiro Yoshita, Evan Fletcher, Oliver Martinez, Mario Ortega,Oliver Martinez, Mario Ortega, Bruce Reed,

Dan Mungas and Charles DeCarli

Department of Neurology and Center for Neuroscience

University of California at Davis

Object & BackgroundObject & Background

How different in Extent and Distribution

(NA)

(MCI)

(AD)

Mild Cognitive Impairment

Normal Aging

White Matter Hyperintensites (WMH)

Alzheimer’s Disease

Subject DemographicsSubject Demographics

Number (F/M) 26 (16/10) 29 (12/17) 33 (23/10)

Age (y) 79.6 ± 6.8* 74.6 ± 8.1 73.4 ± 8.1

range 63-93 60-88 61-91

Education (y) 11.7 ± 5.2 12.6 ± 5.4 10.9 ± 5.8

range 0-18 0-21 0-20

MMSE 19.5 ± 6.8a 26.6 ± 2.3 27.9 ± 2.7

range 0-27 20-30 19-30

Vascular risk score

range

Brain volume (%) 77.3 ± 4.2b 80.7 ± 4.2 83.3 ± 4.3

WMH volume (%) 1.33 ± 1.30c 0.92 ± 0.93 0.56 ± 0.67

Data represented as mean ± standard deviation. * p < 0.01 between AD and NA,a p < 0.001 between AD and others.b p < 0.01 between AD and others, c p < 0.001 between AD and NA.

AD MCI NA

MethodsMethodsImage Analysis

Schema for WMH segmentation and nonlinear transformation for mapping

DeCarli C, et al. Stroke 2005.

A

P

Composite image

A

P

ROI on 3ROI on 3--D WMH mapsD WMH maps

Cg: genu of corpus callosum, Cs: splenium of corpus callosum, Pa: anterior periventricular region, Pb: body of periventricular region, Pp: posterior periventricular region, Po: occipital periventricular region.

Abbreviations:

ROI volume (mm3) :Cg: 263, Cs: 205, Pa: 1325, Pb: 3190, Pp: 2385, Po: 983.

We computed the number of voxel of WMH in each ROI, andcalculated percentage of voxel which have WMH in each ROIof each subject.

(B)

Cg

Cs

Pa

PbPp

Po

(A)

PpPa

Cg

PoPb

Cs

Statistical AnalysisStatistical Analysis

1. Groups were compared using analysis of variance (ANOVA) or Kruskal-Wallis test as appropriate according to the distribution of the data, with post hoc analysis.

2. Correlations were evaluated by Spearman’s ranks correlation test.

3. Results are expressed as mean values ± standard deviation.

4. A p value less than 0.05 was considered significant.

5. Data were analyzed using the Statistical Package for Social Sciences (SPSS for Windows, version 12.0; SPSS Inc, Chicago, IL).

020406080

-20

40 0 -20 -40 -6020 -8060 -100

X= 21(A-1)

020406080

40 0 -20 -40 -6020 -8060 -100

X= 2(A-2)

-20

-60 -20

0

2040

-60-80

-20-40

0 20 40 60-40

Z= 28(B-1)

0

2040

-60-80

-20-40

0 20 40-60 60-20-40

Z= 22(B-2)

020

4060

80

-20-40

0 20 40 60-20-40-60

Y= -4(C-1)

0

20

40

60

80

-20

0 20 40 60-20-40-60Y= -26(C-2)

A

P(D)

WMH frequency maps in all subjectsWMH frequency maps in all subjects

33--D WMH frequency mapsD WMH frequency mapsin each groupin each group

(A) (B) (C)

Three-dimensional reconstruction of the WMH (orange) and ventricular (grey) maps. Orange color indicate the frequency of voxels containing the WMH more than 10%.

(A): group of NA(B): group of MCI(C): group of AD

Group difference of WMH Group difference of WMH severity in each ROIseverity in each ROI

Error bars indicate standard deviation of the mean.

(B): Corpus callosum region

: AD

: MCI

: NA

(%)

0102030405060

Genu Splenium

70*

***

* p < 0.05, ** p < 0.01.

(A): Periventricular region

01020304050607080

Anterior Body Posterior Occipital

(%)

* ***

Correlations between age, MMSE Correlations between age, MMSE and WMH severity in each ROIand WMH severity in each ROI

Pa: anterior periventricular region Pb: body of periventricular region Pp: posterior periventricular region Po: occipital periventricular regionCg: genu of corpus callosum Cs: splenium of corpus callosum

Abbreviations:

AD

Pa Pb Pp Po Cg Cs

0.060 0.068 0.202 0.350 0.120 0.276

0.237 0.355 0.477* 0.553** 0.319 0.055

0.220 0.298 0.411* 0.394* 0.159 0.164

Age

MCI

NA

MMSE

0.355 0.149 0.078 -0.042 0.085 0.032

-0.054 0.018 -0.116 -0.160 0.252 -0.003

0.084 -0.085 -0.051 0.316 0.198 0.192

AD

MCI

NA

* p < 0.05, ** p < 0.01 corrected.

We believe the methods developed here conclusively show the difference of WMH distribution of each cognitive stage.

These observations also support the notion of both a common ischemic etiology and Wallerian degeneration as a mechanism of WMH in each cognitive stage.

Our study shows that measurements of the extent and distribution of WMH in both corpus callosum and periventricular area are suitable for assessment for NA, MCI and AD.

ConclusionConclusion

ThanksThanks

Student:Student:Ryan Berthold,Ryan Berthold,Chris Davies, Chris Davies, Alex Early, Alex Early, Marty Marty GreeniaGreenia, , Evan LloydEvan Lloyd

Staff:Staff:Lauren Lauren CibattariCibattari,,Krista Garcia,Krista Garcia,Oliver Martinez,Oliver Martinez,Mario Ortega,Mario Ortega,BaljeetBaljeet Singh,Singh,Laura Whiteside,Laura Whiteside,Sky Sky RaptentsetsangRaptentsetsang