1. Mouse Model Neuro-Facial Dysmorphology: Translational and Treatment Studies
Charles Goodlett
Bruce Anthony: Sophia
Liang Yun: Huisi Ai
Shiaofen Fang: Yin Liu
Li Shen
Leah Wetherill
Richard Ward
CIFASD 2010 Annual Review

2. Two alcohol doses– 4.8% or 3.6% V/V.
Dose and Stage
Two alcohol doses– 4.8% or 3.6% V/V.
2. Two developmental treatment periods
–E7-E16
–E7-E11
1 week acclimation
Alcohol PF Chow
Either 4.8% or 3.6% Alcohol
or Par fed Diets
From E7-E11 or E7-E17
Mating
Alc-Alc Alc-PF PF-PF Chow E7
through
E11 or E16
Pregnancy treatment same as pre-pregnant
Pre-pregnancy
treatment
All Pups to Surrogate dams by P1
Diets: (PMI) Purina Test Diet (Micro-stabilized Rodent Liquid Diet, LD101A)
with 4.8% and 3.6% alcohol

3. Mouse Nuerofacial dysmorphology
Anthropometic Craniofacial analysis:
Embryonic 17 (Mid gestation equivalent): Microvideo images
Postnatal day 7 &21 (late gestation, toddler): MicroCT
Craniofacial Volume and feature Analysis
Embryo: 2D rendered analysis
Postnatal 7 & 21: MicroCT
Brain volume and neuron analysis
Total and Regional volume, neuronal number

4. Facial Dysmorphology Analysis for E17 Embryos

5. C57BL/6N E17 Embryo Facial Dysmorphology Analysis
Facial morphometric analysis.  P-value of significant treatment main effects (within sub-strain) for each facial measure is indicated. For the paired comparisons, (using Tukey-Kramer procedures) comparisons were made for Alc to Chow, PF to Chow, and Chow to PF groups: Bolded-typed values under columns denote that the overall treatment p-value < the corrected Bonferonni alpha = An asterisk (*) denotes that the Tukey-Kramer pair-wise comparison was significant (p<0.05).  Gray-shaded areas are measurements with significant difference between both control groups and alcohol-exposed group.
Results:
Significant effects specific to alcohol treatment, (i.e., the Alcohol group differed from each respective control group), were identified for two measures of facial depths (upper facial depth; mid-facial depth) in both sub-strains,
Only the B6N strain showed significant effects on two other facial height measures (nasal length; nasal bridge).
The two strains showed markedly different patterns of facial alterations comparing the PF and Chow groups. In particular, the B6N sub-strain showed a significantly greater vulnerability to the pair feeding procedure (in terms of smaller facial measures) than did the B6J sub-strain.

6. C57BL6N Discriminant Analysis of E17 Facial Dysmorphology
Discriminant analysis was done with all 15 measurements for correct predictability. Analysis was performed using backward stepwise model for comparison.  Discriminant analysis was performed using individual animal group with comparisons between Alc and Pair-Fed as well as between Alc to all-non-alcohol litters. Percentage of correct predictability is reported for Alc and Non-Alc groups.  “x” denotes the measurement in each column needed in the final classification model. % Reflects the percentage of correct

7. Silhouette surface extraction Surface alignment
A video 3D image volume was formed by stacking 2D video embryo head images taken from a sequence of angle steps
Silhouette surface extraction
Surface alignment
Feature Selection and Analysis
Validation
CIFASD 2010 Annual Review

8. Posters and Presentations:
Publications:
Shiaofen Fang, Ying Liu, Jeffrey Huang, Sophia Vinci-Booher, Bruce Anthony, and Feng Zhou. Alcohol Exposure Analysis of Mouse Embryos Using Multi-angle Facial Image Analysis. Proc. of ACM Symposium on Applied Computing, 2009,
Bruce Anthony, Sophia Vinci-Booher, Leah Wetherill, Richard. Ward, Charles Goodlett, Feng C. Zhou. Alcohol Induced Facial Dysmorphology in C57BL/6 Mouse Models of Fetal Alcohol Spectrum Disorder. Alcohol and alcoholism, revision under review.
Shiaofen Fang, Ying Liu, Jeffrey Huang, Sophia Vinci-Booher, Bruce Anthony, and Feng Zhou. Surface Feature Analysis using Video Volumes of Mouse Embryos for Fetal Alcohol Syndrome Classification, submitted to International Conference on 3D Data Processing, Visualization, and Transmission (3DPVT).
Yun Lian, Huisi Ai, Bruce Anthony, Leah Wetherill, Charles Goodlett, Feng C. Zhou. (2009) Facial Dysmorphology in Craniofacial Bone Formation in C57BL6 after Alcohol Exposure. In preparation.
Posters and Presentations:
Bruce Anthony, Sophia Vinci-Booher, Leah Wetherill, Richard. Ward, Charles Goodlett,
Feng C. Zhou, Research Society on Alcoholism. Poster presentation. San Diego (2009). Alcohol Induced Facial Dysmorphology in C57BL/6 Mouse Models of Fetal Alcohol Spectrum Disorder.
Yun Lian, Huisi Ai, Bruce Anthony, Leah Wetherill, Charles Goodlett, Feng C. Zhou. Poster presentation Research Society on Alcoholism San Diego (2009). Micro CT Analysis Facial Dysmorphology in Craniofacial Bone Formation in C57BL6 after Alcohol Exposure.
Bruce Anthony. FAS talk. Alcohol Induced Facial Dysmorphology in C57BL/6 Mouse Models of Fetal Alcohol Spectrum Disorder. Research Society on Alcoholism. San Diego (2009).
CIFASD 2010 Annual Review
CIFASD 2010 Annual Review

9. longitudinal 3D-craniofacial dysmorphology: microCT study in the C57BL/6J mouse
Yun Liang1, Huisi Ai1, Li Shen1, Bruce Anthony2,
Richard Ward3, Leah Wetherill4, Feng C. Zhou2
Dept of Radiology1, Dept of Anatomy and Cell Biology2,
Dept of Anthropology3, Dept of Medical and Molecular Genetics4,
IU School of Medicine, Indianapolis, IN
CIFASD 2010 Annual Review

10. Components of MicroCTSystem
X-Ray
Source
Imaging
Volume
Detector
CT System - EVS RS-9
(GE Healthcare, London, Ontario N6G 4X8, Canada)
CIFASD 2010 Annual Review

11. Imaging Techniques Scanner volume coverage: 36.8 mm x 36.8 mm x 78 mm
Scan time: ~ 45 minutes
Recon-resolution: 46 m
Image Quality : HU (noise)
X-ray Dosage: ~ 0.8 Gy
X-ray CT provides detailed bone anatomic information
CIFASD 2010 Annual Review

12. Scan Procedure
CT number Calibration: QA phantom with water, air, acrylic, bone (SB3) inserts
Air
water
Sb3
QA Phantom
CIFASD 2010 Annual Review
CIFASD 2010 Annual Review

13. Image Analysis Methods
Bone Data Analysis:
Part I- 2D craniofacial anthropometric
Part II- individual bone volumes segmented and measured
Image analysis tool: All image analyses are performed with software tool “MxView” (courtesy of Philips Medical Systems) and “Amiral” (Commercial Image Analysis).
CIFASD 2010 Annual Review

14. Table 1. Cranium morphometric measurement and landmarks
Part I
Table 1. Cranium morphometric measurement and landmarks
Type
Number
Measurement
Definition of Landmarks
Width 1
Minimal Frontal Width
Measure distance from junction of Parietal and Frontal bone
suture at the top most inner ridge of the orbital surface 2
Bizygomat Width*
Top view, the distance from the left to right central auditory
canal parallel to a line perpendicular to Frankfort Horizontal 3
Bitragal Width
Measure a straight line distance perpendicular to Frankfort horizontal that begins at the upper most margin of the Tympanic Bulla. 4
Bigonial Width
Measure widest distance between the lateral aspects
of mandible 5
Inner Canthal Width
Measure the shortest distance between the inner ridge of the orbital surface 6
Outer Canthal Width*
(NA)
Measure the longest distance to the inner edge of the
outer orbital socket 7
Philtrum Length*
Distance from the lowest border of the nasal cavity to
the bottom of incisive bone
Depth 8
Upper Facial Depth (L/R)
Distance from the central point of the tympanic bulla
(central ear canal) to the nasion. 9
Mid Facial Depth (L/R)
(central ear canal) to the forward upper tip of the incisive
bone at the junction of the nasal and incisive bones. 10
Lower Facial Depth (L/R)
(central ear canal) to the forward tip of the mandible.
Length 11
Nasal Length
Distance from the Nasion to the forward upper tip of the incisive bone
at the junction of the nasal and incisive bones. 12
Nasal Bridge Length
(Shown in Figure 3i below)
at the junction of the nasal and incisive
bones
New 14
Inner Orbital Width
Front view, the minimum separation (width) between orbital bones. 15
Head Circumference
The outer perimeter of skull enclosure in the axial plane at the level of “nasion” 16
Bony Mandible
Shown in Figure o).
Bony mandible length is the distance between the gonia and the tip of
the menton.
CIFASD 2010 Annual Review

15. Perform anthropometric measurement with landmarks
MicroCT of P21 c57BL mouse facial bone with embryonic analogue division ad landmark measurements.
CIFASD 2010 Annual Review
CIFASD 2010 Annual Review

16. Minimum Frontal Width Upper Facial Depth P7 P21 P7 P21 Nasal Length
CIFASD 2010 Annual Review

17. Sample
ALC-ALC
ALC-PF
PF-PF
CHOW P7 29 35
P21 26 28 40

18. Statistical Analysis for P7
Table 2. Craniofacial morphometric analysis.  P-value of significant treatment main effects for each morphometric measure is indicated. The yellow shaded values under the first column denote that the overall treatment p-value < the corrected Bonferonni alpha= Yellow- shaded areas in the rest of columns denote that the Tukey-Kramer pair- wise comparison was significant (p<0.05). 
CIFASD 2010 Annual Review

19. Dysmophology Classification for P7
CIFASD 2010 Annual Review

20. Dysmophology Classification for P21
CIFASD 2010 Annual Review

21. Future Work
Perform longitudinal comparison for P7 and P21 at 4.8% alcohol;
Continue examination for 3.6% alcohol liquid diet;
Correlate facial (video) facial and craniofacial dysmorphology.
CIFASD 2010 Annual Review

22. Part II. 3D Bone Analysis
The bone growth in differentiated anthropometry regions are evaluated.
The development of ossification centers and sutures are being evaluated.
The Shape analysis of regions and areas are planned.
CIFASD 2010 Annual Review

23. Procrustes Analysis
Definition: A means of comparing two arbitrary multidimensional configurations by removing the transnational and rotational differences between them such that they best fit.
Procedures: Using anthropometric landmarks to match the overall sizes of the objects, and to identify and compare the regional/global differences (in shape, size,…).
CIFASD 2010 Annual Review

24. Landmarks Landmarks Index Name 1 forward center tip of nasal bone fctn2
nasion n
3l, 3r
3,4
inner canthal width
icw
4l, 4r
5,6
minimal frontal width
mfw
5l, 5r
7,8
bigonial width
bgo
6l, 6r
9,10
bitragal width
btr 7 11
forward tip of mandible
ftm
8l, 8r
12,13
nasal length nl
9l, 9r
14,15
interior point of orbital socket in
10l, 10r
16,17
exterior point of orbital socket ex
11l, 11r
18,19
bony mandible
man 12 20
bregma
breg
CIFASD 2010 Annual Review

25. (1)
(8)
(2)
(3)
(4)
(9)
(10)
CIFASD 2010 Annual Review

26. CIFASD 2010 Annual Review (1) (2) (3) (4) (11) (8) (9) (10) (5) (6)
(7)
CIFASD 2010 Annual Review

27. Bone Segmentation: P21 Examples
Frontal
Occipital
Parietal
Jaw
CHOW
ALC-ALC
ALC-PF

28. Sample
ALC-ALC
ALC-PF
CHOW P7 19 10 32
P21 20 36

29. Bone Volumes vs. Centroid Size
The Centroid is defined as the geometric center of all landmarks;
The Centroid_size is defined as the square root of the summed squared distances between the landmarks and the centroid.
CIFASD 2010 Annual Review

30. General Linear Model (GLM) on P21

32. GLM on P21 without Covariates
Tests of Between-Subjects Effects
Source
Dependent Variable df F
Sig.
Frontal 2
10.715
.000
Parietal
25.773
Occipital
19.507
Jaw
4.264
.018
Centroid_Size
16.464
Pairwise Comparisons (Sig.)
ALC-ALC vs ALC-PF
ALC-ALC vs CHOW
ALC-PF vs CHOW
Frontal
.003
.000
.296
Parietal
.889
Occipital
.042
Jaw
.110
.005
.281
Centroid_Size
.712

33. GLM on P21 with Centroid_Size as Covariate
Tests of Between-Subjects Effects
Source
Dependent Variable df F
Sig.
Corrected Model
Frontal 3
47.157
.000
Parietal
52.713
Occipital
45.181
Jaw
21.676
Group 2
.705
.498
7.004
.002
5.413
.006
1.178
.314
Pairwise Comparisons (Sig.)
ALC-ALC vs ALC-PF
ALC-ALC vs CHOW
ALC-PF vs CHOW
Frontal
.723
.553
.248
Parietal
.001
.595
Occipital
.472
.006
.016
Jaw
.146
.512
.275

34. Summary ALC-ALC vs CHOW: easy to distinguish
P21: 7 out of 9 measures can tell the difference
P7: 5 out of 9 measures can tell the difference
ALC-ALC vs ALC-PF: easy to distinguish
P21: 5 out of 9 measures can tell the difference
P7: 2 out of 9 measures can tell the difference
ALC-PF vs CHOW: very hard to distinguish
Only Occipital at P21 can tell the difference (p=0.42, p=0.16)

35. Summary Interesting measures
Centroid_Size at both P7 and P21 stages can distinguish ALC_ALC from ALC_PF and CHOW, but not between ALC_PF and CHOW.
Occipital at P21 can distinguish all three groups from one another

36. CIFASD 2010 Annual Review