ORIGINAL ARTICLE

Usefulness of Fetal Three-Dimensional Ultrasonographyfor Detecting of Congenital Heart Defects and AssociatedSyndromes

Nadia Zabadneh • Claudia Santagati •

Elena Reffo • Roberta Biffanti • Alessia Cerutti •

Nicola Maschietto • Ornella Milanesi

Received: 4 October 2010 / Accepted: 22 March 2011 / Published online: 9 April 2011

� Springer Science+Business Media, LLC 2011

Abstract Congenital heart defects (CHDs) occur in 1%

of live-born infants and frequently are associated with

extracardiac malformations. This study aimed to assess the

feasibility and accuracy of three-dimensional ultrasonog-

raphy (3DUS) in fetuses with CHD and to investigate

whether 3DUS can add information about the heart and

general fetal morphology that shows other congenital

malformations or suggests syndromes. For 30 fetuses

affected by CHD, 3DUS was performed using a Sonos

7500 ultrasound machine with a cardiac 3D transducer. In

44% of the exams, 3DUS was completely diagnostic for the

CHD, providing additional information in 28% of the

exams. Furthermore, 3DUS showed 82% of associated

malformations, providing the complete diagnosis in 57% of

the cases and helping with recognition of syndromes in

others. The diagnostic accuracy of 3DUS was superior,

with a higher number of acquisitions per exam. Perfor-

mance was better in fetuses younger than 24 weeks for

general morphologic details and in fetuses older than

24 weeks for the heart morphology.

Keywords Congenital � 3D Echocardiography �Heart defects � Fetus � Prenatal diagnosis � Syndromes

Introduction

Congenital heart defects occur frequently in malformation

syndromes. Approximately 70% of spontaneous aborted

and stillborn fetuses and 25% of infants with a congenital

heart defect (CHD) have associated extracardiac malfor-

mations. Detection of a heart defect should prompt a

detailed search for associated extracardiac malformations.

Parents expect a complete and accurate prenatal diagnosis

and appropriate information regarding the natural history

of the cardiac defect and any associated abnormality [6].

Two-dimensional (2D) ultrasonography imaging is the

principal method for obtaining diagnostic information

about the fetus, but proper interpretation depends on cre-

ating a mental three-dimensional (3D) reconstruction from

the images [4]. The mental process of converting 2D into

3D images depends on the skill, training, and experience of

the operator [20], as evidenced by the wide variability in

diagnostic accuracy for detecting congenital anomalies [12,

19]. In the last few years, 3D imaging techniques have been

applied in fetal ultrasonography with rapid progress [1, 3,

5, 7–11, 14–16, 23, 24, 27, 28], with hope that 3D images

would reduce operator dependency and improve diagnostic

accuracy.

This study aimed to assess the feasibility and accuracy

of 3D ultrasonography in a group of fetuses with a CHD

diagnosed by 2D ultrasonography and to assess the specific

role of the 3D reconstruction in the current practice of fetal

cardiology. Specifically, we aimed to determine whether

3D images added information about the heart and general

fetal morphology that would disclose other congenital

malformations or suggest syndromes, yielding a more

complete diagnosis. To our knowledge, this is the first

report describing this matter from the pediatric cardiology

perspective.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00246-011-9977-9) contains supplementarymaterial, which is available to authorized users.

N. Zabadneh � C. Santagati � E. Reffo � R. Biffanti �A. Cerutti � N. Maschietto � O. Milanesi (&)

Department of Pediatrics, University of Padova,

Via Giustiniani 3, 35128 Padova, Italy

e-mail: ornella.milanesi@unipd.it

123

Pediatr Cardiol (2011) 32:724–736

DOI 10.1007/s00246-011-9977-9

Materials and Methods

Patients

All fetuses with congenital heart anomalies diagnosed by

2D echocardiography at the Fetal Cardiac Unit of our

Department from October 2008 through February 2009

were eligible for enrollment in the study. Subjects were

enrolled if the parents agreed and gave informed consent.

The 3D ultrasonography study included one or more

real-time or volumetric 3D data sets of the heart, head,

face, limbs, and genitalia. If the patient underwent a fol-

low-up 2D exam, a repeat 3D exam was performed as well

if the parents consented. The outcome of each pregnancy

was determined. For pregnancies that went to term, follow-

up data were obtained from the medical record of the

infant. The fetopsy report was reviewed for all pregnancies

ending in spontaneous or planned termination.

The research protocol was approved by the Hospital

Committee on Clinical Investigation, and the procedures

followed were in accordance with institutional guidelines

for clinical research studies and protection of patient

confidentiality.

Table 1 Cardiac defects in 30 enrolled fetuses with congenital heart

defect (CHD)

Pathologies No. observed in 30

fetuses

%

Ventricular septal defect 11 20.8

Aortic arch hypoplasia 5 9.4

Pulmonary stenosis 4 7.5

Hypoplastic left heart syndrome 3 5.7

Abnormal systemic venous return 2 3.8

Complete atrioventricular septal

defect

2 3.8

Partial atrioventricular septal defect 2 3.8

Univentricular heart 2 3.8

Total anomalous pulmonary venous

return

2 3.8

Cardiac tumors 2 3.8

Heterotaxy 2 3.8

Tetralogy of Fallot 2 3.8

Transposition of the great arteries 2 3.8

Ebstein anomaly 1 1.9

Tricuspid atresia 1 1.9

Pulmonary atresia 1 1.9

Dilated cardiomyopathy 1 1.9

Aortic coarctation 1 1.9

Dextrocardia 1 1.9

Atrial septal defect 1 1.9

Mitral valve regurgitation 1 1.9

Tricuspid valve hypoplasia 1 1.9

Apex malposition 1 1.9

Truncus arteriosus 1 1.9

Left ventricular-aortic tunnel 1 1.9

Total 53 100

Table 2 Morphologic defects in 30 enrolled fetuses with congenital

heart defect (CHD)

Pathologies No. observed

in 30 fetuses

%

Dysmorphic face 4 14.3

Low-set ears 3 10.7

Micrognathia 3 10.7

Cleft lip-palate 2 7.1

Cranial suture diastasis 2 7.1

Mongoloid face 2 7.1

Prognathia 2 7.1

Brachycephaly 1 3.6

Face suggestive of catch 22 1 3.6

Clitoral hypertrophy 1 3.6

Labia majora hypertrophy 1 3.6

Nasal bone hypoplasia 1 3.6

Phalanx agenesia 1 3.6

Prominent occiput 1 3.6

Small feet 1 3.6

Small hands 1 3.6

Ulna’s agenesia 1 3.6

Total 28 100

Table 3 Visualization rate and image quality for each cardiac segment in 42 exams

Cardiotoracic

ratio n (%)

Visceral

atrial situs

n (%)

Cardiac

apex

n (%)

Systemic

veins

n (%)

Pulmonary

veins n (%)

Atrioventricular

connection

n (%)

Ventricular

morphology

n (%)

Septum

n (%)

Outflow

tracts

n (%)

Great

arteries

n (%)

Aortic

arch

n (%)

Visualized 32 (76) 23 (55) 39 (93) 25 (60) 8 (19) 31 (74) 33 (79) 34 (81) 33 (79) 31 (74) 9 (21)

Not visualized 10 (24) 19 (45) 3 (7) 17 (40) 34 (81) 11 (26) 9 (21) 8 (19) 9 (21) 11 (26) 33 (79)

Diagnostic 23 (55) 18 (43) 31 (74) 14 (33) 2 (5) 19 (45) 21 (50) 18 (43) 19 (45) 15 (36) 3 (7)

Not diagnostic 19 (45) 24 (57) 11 (26) 28 (67) 40 (95) 23 (55) 21 (50) 24 (57) 23 (55) 27 (64) 39 (93)

Pediatr Cardiol (2011) 32:724–736 725

123

Instrumentation and Data Acquisition

Each exam was performed transabdominally using a Sonos

7500 ultrasound machine (Philips Medical Systems, Bot-

hell, WA, USA) with a cardiac 2D matrix phased array

2–4 MHz transducer (94, Sonos 7500; Philips Medical

Systems) designed for real-time 3D cardiac imaging.

Images were acquired in live 3D mode as a 60� 9 30�pyramid or in full-volume acquisition mode as a 60� 9 60�pyramid [13, 17] . In the latter mode, the data set was

acquired as four 60� 9 15� slices from four different car-

diac cycles and gated to the mother’s electrocardiogram

(ECG).

Images of the heart, head, face, limbs, and genitalia

were obtained at medium- and high-line density, frequency

fusion of 3, gain and compression at midlevel settings,

power output at 0, and time gain compensation (TGC) set

to produce faint intracavitary blood pool echoes.

Offline Processing

Volumetric data sets were stored and evaluated offline by a

pediatric cardiologist with 3D diagnostic imaging experi-

ence. The stored data were manipulated by adjusting gain

and compression, smoothing settings, and using cutting

planes to expose anatomic features.

Data Analysis

The feasibility study investigated the frequency with which

specific cardiac anatomic features and general fetal mor-

phologic features could be displayed in the 3D data set and

the diagnostic quality of the images. First, the fetal cardiac

and overall fetal anatomies were divided into several

domains. For the heart, these included 11 segments:

cardiothoracic ratio, visceral-atrial situs, location of the

cardiac apex, systemic veins, pulmonary veins, atrioven-

tricular connection, ventricles, septum, outflow tracts, great

arteries, and aortic arch. For the general fetal anatomy, we

considered the skull, face, lips, palate, hands, upper limbs,

feet, lower limbs, and genitalia.

Next, the data sets were prospectively examined by a

blinded reviewer (pediatric cardiologist experienced in 3D

fetal imaging) to determine how often the anatomic fea-

tures of each domain were displayed in the 3D data sets

(structure seen or not seen) as well as the diagnostic quality

of the images (seen but not diagnostic). Finally, as com-

plete a diagnosis as possible was formulated from the

review.

The results of the blinded review were compared with

the 2D fetal echocardiogram report to determine what

additional information was provided by the 3D imaging.

The accuracy study involved comparing the diagnoses

based on both the blinded review of the 3D data sets and

the 2D data with the postnatal or fetopsy findings.

Statistical Analysis

Data are presented as raw data and percentages. The sig-

nificance of differences in proportions between groups was

Fig. 1 Examples of normal cardiac anatomy. a Aortic arch. b Long-

axis-like view in a normal heart. c Caval veins and atria

726 Pediatr Cardiol (2011) 32:724–736

123

Ta

ble

4C

om

par

iso

no

fca

rdia

cd

iag

no

ses

infe

tal

thre

e-d

imen

sio

nal

(3D

),fe

tal

2D

,an

dp

ost

nat

al/f

eto

psy

fin

din

gs

No

.o

f

exam

No

.o

f

case

Ges

tati

on

al

wee

k

Dia

gn

ost

icfi

nd

ing

s

3D

Fet

alec

ho

card

iog

rap

hy

2D

Fet

alec

ho

card

iog

rap

hy

Po

stn

atal

/fet

op

sy

Fin

din

gs

11

24

VS

D,

ven

tric

ula

r-ao

rtic

tun

nel

,ao

rtic

atre

sia

VS

D,

ven

tric

ula

r-ao

rtic

tun

nel

VS

D,

ven

tric

ula

r-ao

rtic

tun

nel

,ao

rtic

atre

sia

21

a3

0V

SD

VS

D,

tun

nel

VS

D,

ven

tric

ula

r-ao

rtic

tun

nel

,ao

rtic

atre

sia

32

30

Dis

pro

po

rtio

nao

rta\

pu

lmo

nar

yar

teri

esH

yp

op

last

icao

rtic

arch

Co

arct

atio

n

43

37

Dis

pro

po

rtio

nR

V[

LV

Co

arct

atio

nH

yp

op

last

icao

rtic

arch

54

37

Co

mm

on

AV

SD

and

TO

F,

bal

ance

dv

entr

icle

s,

pu

lmo

nar

yv

alv

est

eno

sis

AV

SD

and

TO

FC

om

mo

nA

VS

Dan

dT

OF

,b

alan

ced

ven

tric

les,

pu

lmo

nar

yv

alv

est

eno

sis

65

21

VS

Dap

ical

,T

OF

,H

yp

op

last

icp

ulm

on

ary

val

ve

TO

FT

OF

,m

ild

pu

lmo

nar

yv

alv

est

eno

sis

75

a2

2N

ET

OF

,m

ild

pu

lmo

nar

yv

alv

est

eno

sis

TO

F,

mil

dp

ulm

on

ary

val

ve

sten

osi

s

86

32

NE

Ao

rtic

val

ve

sten

osi

sH

yp

op

last

icao

rtic

arch

97

31

Co

mp

lete

TG

AC

om

ple

teT

GA

and

larg

eV

SD

Co

mp

lete

TG

Aan

dla

rge

VS

D

10

82

3N

EM

usc

ula

rV

SD

Mu

scu

lar

VS

D

11

8a

30

Mu

scu

lar

VS

DM

usc

ula

rV

SD

Mu

scu

lar

VS

D

12

92

1H

LH

S,

VS

DH

LH

S,

VS

DH

LH

S,

VS

D

13

10

22

NE

HL

HS

No

tav

aila

ble

14

11

23

NE

Rig

ht-

sid

edst

om

ach

,A

VS

D,

RV

giv

ing

rise

on

lyto

aort

icar

ch,

pu

lmo

nar

yat

resi

a

Het

ero

tax

y,

lev

oca

rdia

,b

ilat

eral

SC

Vs,

com

mo

n

AV

val

ve,

2p

ost

erio

rV

SD

can

alty

pe,

pu

lmo

nar

yat

resi

a

15

11

a3

0B

ilat

eral

SC

Vs,

AV

SD

,R

Vg

ivin

gri

seo

nly

to

aort

icar

ch,

pu

lmo

nar

yat

resi

a,

Sit

us

inv

ersu

s,A

VS

D,

RV

giv

ing

rise

on

lyto

aort

icar

ch,

pu

lmo

nar

yat

resi

a

Het

ero

tax

y,

lev

oca

rdia

,b

ilat

eral

SC

Vs,

Co

mm

on

AV

val

ve,

2p

ost

erio

rV

SD

can

alty

pe,

pu

lmo

nar

yat

resi

a

16

12

32

Rh

abd

om

yo

mas

,in

flo

w-o

utfl

ow

no

to

bst

ruct

edR

hab

do

my

om

as,

infl

ow

-ou

tflo

wn

ot

ob

stru

cted

Rh

abd

om

yo

mas

,in

flo

w-o

utfl

ow

no

to

bst

ruct

ed

17

12

a3

7R

hab

do

my

om

as,

infl

ow

-ou

tflo

wn

ot

ob

stru

cted

Rh

abd

om

yo

mas

,in

flo

w-o

utfl

ow

no

to

bst

ruct

edR

hab

do

my

om

as,

infl

ow

-ou

tflo

wn

ot

ob

stru

cted

18

13

20

Sin

gle

ven

tric

leh

eart

Sin

gle

ven

tric

leh

eart

,su

bp

ulm

on

ary

sten

osi

sN

ot

avai

lab

le

19

14

24

HL

HS

,sm

all

aort

icar

ch,

pu

lmo

nar

yv

ein

sn

ot

seen

Hy

po

pla

stic

mit

ral

val

ve,

HL

HS

,h

yp

op

last

ic

asce

nd

ing

aort

icar

ch.

TP

AV

C

HL

HS

,ao

rtic

val

ve

atre

sia,

hy

po

pla

stic

asce

nd

ing

aort

icar

ch

20

14

a3

0H

LH

S,

smal

lao

rtic

arch

,p

ulm

on

ary

vei

ns

no

t

seen

Hy

po

pla

stic

mit

ral

val

ve,

HL

HS

,h

yp

op

last

ic

asce

nd

ing

aort

icar

chT

PA

VC

HL

HS

,ao

rtic

val

ve

atre

sia,

hy

po

pla

stic

asce

nd

ing

aort

icar

ch

21

15

19

TG

A,

VS

DH

yp

op

last

icm

itra

lv

alv

e,h

yp

op

last

icL

V,

TG

A,

VS

D

Hy

po

pla

stic

LV

,T

GA

,V

SD

22

16

24

Ap

exm

alp

osi

tio

nA

pex

mal

po

siti

on

No

tav

aila

ble

23

17

20

Sm

all

VS

DS

mal

lV

SD

Sm

all

VS

D

24

18

20

VS

DE

bst

ein

ano

mal

yE

bst

ein

ano

mal

y

25

18

a2

7M

ild

Eb

stei

nan

om

aly

,n

oV

SD

Eb

stei

nan

om

aly

Mil

dE

bst

ein

ano

mal

y

26

19

32

AV

SD

par

tial

/in

term

ediu

m,

2d

iffe

ren

tv

alv

e

AV

,p

rob

able

mit

ral-

clef

t,A

SD

pri

mu

m

AV

SD

par

tial

AV

SD

par

tial

/in

term

ediu

m,

2d

iffe

ren

tv

alv

e

AV

,p

rob

able

mit

ral-

clef

t,A

SD

pri

mu

m

Pediatr Cardiol (2011) 32:724–736 727

123

Ta

ble

4co

nti

nu

ed

No

.o

f

exam

No

.o

f

case

Ges

tati

on

al

wee

k

Dia

gn

ost

icfi

nd

ing

s

3D

Fet

alec

ho

card

iog

rap

hy

2D

Fet

alec

ho

card

iog

rap

hy

Po

stn

atal

/fet

op

sy

Fin

din

gs

27

20

16

Dex

tro

card

iaH

yp

op

last

icao

rta

and

LV

,H

yp

op

last

icm

itra

l

val

ve

Dex

tro

card

ia,

inte

rru

pte

dIC

V,

pat

ent

left

SC

V,

abse

nt

rig

ht

VC

S,

TP

AV

C,

AV

SD

,d

ysp

last

ic

pu

lmo

nar

yv

alv

e,ap

ical

VS

D

28

20

a2

9D

extr

oca

rdia

,p

aten

tle

ftS

CV

,A

VS

DS

itu

sso

litu

s/am

big

uo

us,

inte

rru

pte

dIC

V,

Par

tial

AV

PC

,ri

gh

td

om

inan

tA

VS

D,

dy

spla

stic

pu

lmo

nar

yv

alv

e

Dex

tro

card

ia,

inte

rru

pte

dIC

V,

pat

ent

left

SC

V,

abse

nt

rig

ht

VC

S,

TP

AV

C,

AV

SD

,d

ysp

last

ic

pu

lmo

nar

yv

alv

e,ap

ical

VS

D

29

21

26

Inte

rru

pte

dao

rtic

arch

Hy

po

pla

stic

aort

icar

ch,

VS

DH

yp

op

last

icao

rtic

arch

,la

rge

VS

D

30

22

23

NE

Sm

all

VS

D,

hy

po

pla

stic

aort

icar

ch/c

oar

ctat

ion

TP

AV

C,

hy

po

pla

stic

tran

sver

seao

rtic

arch

31

22

a2

8N

E,

sug

ges

tiv

eo

fT

PA

VC

Sm

all

VS

D,

hy

po

pla

stic

aort

icar

ch/c

oar

ctat

ion

TP

AV

C,

hy

po

pla

stic

tran

sver

seao

rtic

arch

32

23

26

Po

ster

ior

VS

D,

can

alty

pe

Po

ster

ior

VS

D,

can

alty

pe

Po

ster

ior

VS

D,

can

alty

pe

33

23

a3

4P

ost

erio

rV

SD

,ca

nal

typ

eP

ost

erio

rV

SD

,ca

nal

typ

eP

ost

erio

rV

SD

,ca

nal

typ

e

34

24

28

Su

bp

ulm

on

ary

VS

DS

ub

pu

lmo

nar

yV

SD

Su

bp

ulm

on

ary

VS

D

35

25

22

Tri

cusp

idat

resi

aT

ricu

spid

atre

sia

Tri

cusp

idat

resi

a

36

25

a3

2T

ricu

spid

atre

sia,

bu

lbo

ven

tric

ula

rfo

ram

en,

seco

nd

ary

cham

ber

Tri

cusp

idat

resi

aT

ricu

spid

atre

sia

37

26

21

Sin

gle

ven

tric

leh

eart

,h

yp

op

last

ictr

icu

spid

val

ve

Tri

cusp

idat

resi

aS

ing

lev

entr

icle

hea

rt,

hy

po

pla

stic

tric

usp

id

val

ve

38

26

a3

0S

ing

lev

entr

icle

hea

rt,

hy

po

pla

stic

tric

usp

id

val

ve

Sin

gle

ven

tric

leh

eart

,h

yp

op

last

ictr

icu

spid

val

ve

Sin

gle

ven

tric

leh

eart

,h

yp

op

last

ictr

icu

spid

val

ve

39

27

30

VS

Ds

Mit

ral

val

ve

reg

urg

itat

ion

,p

ulm

on

ary

val

ve

sten

osi

s

Mit

ral

val

ve

reg

urg

itat

ion

,p

ulm

on

ary

val

ve

sten

osi

s

40

28

32

Car

dio

meg

aly

Car

dio

meg

aly

Car

dio

meg

aly

41

29

21

Tru

ncu

sar

teri

osu

sty

pe

1T

run

cus

arte

rio

sus

Tru

ncu

sar

teri

osu

sty

pe

I

42

30

32

Rh

abd

om

yo

mas

Rh

abd

om

yo

mas

,in

flo

w-o

utfl

ow

trac

tn

ot

ob

stru

cted

Rh

abd

om

yo

mas

,in

flo

w-o

utfl

ow

trac

tn

ot

ob

stru

cted

VS

Dv

entr

icu

lar

sep

tal

def

ect,

LV

left

ven

tric

le,

RV

rig

ht

ven

tric

le,

AV

SD

atri

ov

entr

icu

lar

sep

tal

def

ect,

TO

Fte

tral

og

yo

fF

allo

t,N

Ed

iag

no

stic

hy

po

thes

isn

ot

exp

ress

ible

,T

GA

tran

spo

siti

on

of

the

gre

atar

teri

es,

HL

HS

hy

po

pla

stic

left

hea

rtsy

nd

rom

e,S

CV

sup

erio

rca

val

vei

n,

TP

AV

Cto

tal

pu

lmo

nar

yan

om

alo

us

vei

nco

llec

t,A

SD

atri

alse

pta

ld

efec

t,IC

Vin

feri

or

cav

alv

ein

aF

oll

ow

-up

exam

inat

ion

728 Pediatr Cardiol (2011) 32:724–736

123

tested using the chi-square test or Fisher’s test. A p value

less than 0.05 was considered statistically significant.

Results

Patients

Of the 230 fetuses scanned between October 2008 and

February 2009, 35 (15%) had a CHD diagnosed by 2D fetal

echocardiography. The gestational ages ranged from 16 to

38 weeks (mean, 24.4 weeks; median, 23 weeks). Of the

35 parents, 30 gave consent for the 3D ultrasonography

evaluation and were enrolled in the study. We performed

12 follow-up exams, for a total of 42 examinations in 30

fetuses. The number of acquisitions ranged from 1 to 6

(average, 2.5; median, 2) for the cardiac domains and from

1 to 15 (average, 5; median, 5) for the fetal morphologic

domains. The 3D acquisitions required an average of 4 min

(range, 3–6 min), whereas the postprocessing and analysis

required a longer time, 20–55 min (average, 35 min).

The outcome of the pregnancies was 24 live-born infants

and six voluntary terminations of pregnancy. Postnatal or

fetopsy data were available for 23 of the 24 neonates and

for four of the six abortions. These data showed a wide

spectrum of cardiac defects, with some fetuses having more

than one defect (Table 1). Nine fetuses had a total of 28

extracardiac abnormalities, suggesting a syndrome in eight

fetuses (Table 2).

Heart Segments

The rate for visualization of the 11 cardiac segments was

highly variable, ranging from 19 to 93% (Table 3). Fig-

ure 1 presents examples of normal cardiac anatomy. The

aortic arch (21%) and the pulmonary veins (19%) were the

least consistently imaged, whereas the atrioventricular

canal, ventricles, outflows, and great arteries were imaged

in 74–93% of cases. Diagnostic quality images were

obtained in fewer segments (5–55%) (Table 3), but the rate

of obtaining diagnostic images was higher if at least three

acquisitions were performed during the exam (p \ 0.01;

p value, 0.000008). Images were more likely to be diag-

nostic after 24 weeks of gestation (p \ 0.01; p value,

0.00007) due to the intrinsic resolution of the 3D matrix

probe and the smaller dimension of the heart in younger

fetuses.

Compared with postnatal or fetopsy findings, the 3D

echo was completely diagnostic in 17 exams (44%), par-

tially diagnostic in 14 exams (36%), and not diagnostic in 8

exams (20%). As analyzed by patient, combining all the

exams performed for a single patient, the 3D exam was

completely diagnostic in 52%, partially diagnostic in 37%,

and nondiagnostic in 11% of cases (see Table 4 compari-

son of 2D and 3D postnatal findings). Compared with the

2D ultrasonography, the 3D exam provided additional

information about the cardiac defects in 28% of cases.

Fetal Morphology

Again, substantial variability was noted among the fetal

morphology domains for visualization, ranging from 25%

for the feet to 86% for the face (Table 5). See the examples

of normal morphologic features in Fig. 2. Diagnostic

images were obtained in slightly fewer domains: 18% for

the feet to 60% for the skull (Table 5). In general, the face,

lip, palate, and skull were more often imaged than the

limbs and genitalia, and the images were more often

diagnostic, probably due to the lower number of specific

acquisitions for these latter structures. More acquisitions

per exam and gestational age younger than 24 weeks were

associated with a higher likelihood of obtaining diagnostic

images (p value, 0.000015 for gestational weeks fewer than

24; p value, 0.000035 for more than 4 acquisitions). In

contrast to the heart, the somatic anatomy was better

imaged in younger fetuses, whose bodies were more

completely enclosed in the 3D sector.

The postnatal or fetopsy findings disclosed 28 extra-

cardiac malformations in nine infants or aborted fetuses

(Table 2). The 3D images suggested that multiple malfor-

mations were present in eight fetuses of the nine cases and

that a syndrome likely was present in seven cases (in

Table 6 see comparison of 2D and 3D postnatal findings).

Images of the malformation were obtained for 23 (82%) of

Table 5 Visualization rate and image quality for each morphologic segment in 42 exams

Skull

n (%)

Face

n (%)

Lips and

palate n (%)

Hands

n (%)

Superior limbs

n (%)

Feet

n (%)

Inferior limbs

n (%)

Genitalia

n (%)

Visualized 29 (69) 36 (86) 34 (81) 22.5 (54) 17.5 (42) 10.5 (25) 17.5 (42) 22 (52)

Not visualized 13 (31) 6 (14) 8 (19) 19.5 (46) 24.5 (58) 31.5 (75) 24.5 (58) 20 (48)

Diagnostic 25 (60) 22 (52) 28 (67) 11.5 (27) 15 (36) 7.5 (18) 14.5 (35) 20 (48)

Not diagnostic 17 (40) 20 (48) 14 (33) 30.5 (73) 27 (64) 34.5 (82) 27.5 (65) 22 (52)

a Considering limbs, feet and hands, we counted 0.5 for each single segment

Pediatr Cardiol (2011) 32:724–736 729

123

the 28 anomalies present. A complete diagnosis was pos-

sible for 16 malformations (57%), a partial diagnosis for

seven malformations (25%), and no diagnosis in three

cases (11%). Two anomalies (7%) were misdiagnosed: a

fetus with hypertrophy of the labia majora was diagnosed

as normal, and clinodactyly was incorrectly diagnosed in a

fetus with small hands at birth. In general, skull anomalies

(micrognathia, prognathia, brachycephaly, cranial suture

diastase), facial anomalies (low-set ears, dysmorphic face),

and cleft lips and palate were more readily detected than

abnormalities of the limbs or genitalia. Figure 3 shows a

cleft lip diagnosed with 3D examination. More examples of

diagnostic findings are discussed later.

Discussion and Conclusions

This study demonstrated that 3D ultrasonography can

diagnose many cardiac defects, and in nearly one-third

Fig. 2 Normal fetal

morphology visualized with

three-dimensional

ultrasonography (3DUS).

a, b Face. c Hands. d Female

genitalia. e Male genitalia

730 Pediatr Cardiol (2011) 32:724–736

123

Ta

ble

6C

om

par

ing

extr

acar

dia

cd

iag

no

ses

infe

tal

thre

e-d

imen

sio

nal

(3D

)an

dp

ost

nat

al/p

ost

abo

rtio

nfi

nd

ing

s

No

.o

f

exam

No

.o

f

case

Ges

tati

on

al

wee

k

Mal

form

ed

fetu

s

Dia

gn

ost

icfi

nd

ing

s

Ind

icat

ion

tofe

tal

ech

oca

rdio

gra

ph

y3

DF

etal

dia

gn

osi

sP

ost

nat

al/p

ost

abo

rtio

n

54

37

1S

usp

ect

CH

DN

EM

on

go

loid

face

,n

uch

alfo

ld,

AS

D,

TO

F:

Do

wn

syn

dro

me

6,

75

21

,2

22

Su

spec

tC

HD

Fac

esu

gg

esti

ve

of

mic

rod

elet

ion

22

q1

1F

ace

sug

ges

tiv

eo

fm

icro

del

etio

n2

2q

11

,T

OF

10

,1

18

23

–3

03

Su

spec

tC

HD

,m

icro

gn

ath

ia,

nu

chal

cyst

ich

yg

rom

a,IU

GR

Mic

rog

nat

hia

,d

iast

atic

sag

itta

lsu

ture

,

clin

od

acty

ly

Mic

rog

nat

hia

,d

iast

atic

sag

itta

lan

dm

eto

pic

sutu

re,

mic

roce

ph

aly

,sy

no

ph

rys,

hy

per

tric

ho

sis,

low

-set

ear,

smal

lfe

etan

d

han

ds,

VS

D:

Co

rnel

iad

eL

ang

esy

nd

rom

e

12

92

14

Su

spec

tC

HD

Pro

gn

ath

ia,

flat

nas

alro

ot,

lon

gan

dth

in

no

sefi

lter

,lo

w-s

etea

rs

Dy

smo

rph

icfa

ce,

low

-set

ears

,T

GA

,V

SD

21

15

19

5S

usp

ect

CH

D,m

icro

gn

ath

ia,ag

enes

iao

fa

fore

arm

seg

men

t

Mic

rog

nat

hia

,ag

enes

iao

fa

fore

arm

seg

men

t,an

om

alo

us

han

d

Mic

rog

nat

hia

,lo

w-s

etea

rs,

agen

esia

of

1

ph

alan

xo

fle

fth

and

,ag

enes

iao

fle

ftu

lna,

hy

po

pla

sia

of

left

rad

ius,

hy

po

pla

sia

LV

,T

GA

,

VS

D

23

17

20

?3

6:N

T3

,37

,si

ng

leu

mb

ilic

alar

tery

,

sto

mac

hn

ot

vis

ual

ized

Mic

rog

nat

hia

,b

rach

yce

ph

aly

,p

rom

inen

t

occ

ipu

t,d

iast

atic

sag

itta

lsu

ture

,cl

eft

lip

Dy

smo

rph

icsk

ull

and

face

,cl

eft

lip

,

hy

po

telo

rism

,re

nal

ano

mal

ies,

VS

D:

Ed

war

ds

syn

dro

me

30

,3

12

22

3–

28

7S

usp

ect

CH

DD

ysm

orp

hic

face

,p

rog

nat

hia

,an

om

alo

us

gen

ital

ia

Mic

rop

hth

alm

ia,

sun

ken

eyes

,h

yp

ertr

ich

osi

s,

pro

gn

ath

ia,

hy

per

tro

ph

yo

fth

ecl

ito

ris,

TP

AV

C:

Tra

nsl

oca

tio

nt(

4:1

5)

32

,3

32

32

6–

34

8S

usp

ect

CH

DH

yp

op

last

icn

asal

bo

ne,

up

war

dsl

ant

to

the

eye,

epic

anth

icfo

ldo

fth

eey

elid

,

ten

t-sh

aped

lip

s

Mo

ng

olo

idfa

ce,

lab

iam

ajo

ra,

hy

per

tro

ph

y,

VS

D:

Do

wn

syn

dro

me

35

,3

62

52

2–

32

9S

usp

ect

CH

DC

left

lip

,cl

eft

pal

ate

Cle

ftli

p,

clef

tp

alat

e,tr

icu

spid

atre

sia

CH

Dco

ng

enit

alh

eard

def

ect,

NE

dia

gn

ost

ich

yp

oth

esis

no

tex

pre

ssib

le,

AV

SD

atri

ov

entr

icu

lar

sep

tal

def

ect,

TO

Fte

tral

og

yo

fF

allo

t,IU

GR

intr

aute

rin

eg

row

thre

stri

ctio

n,

TG

Atr

ansp

osi

tio

n

of

the

gre

atar

teri

es,

VS

Dv

entr

icu

lar

sep

tal

def

ect,

LV

left

ven

tric

le,

TP

AV

Cto

tal

pu

lmo

nar

yan

om

alo

us

vei

nco

llec

t

Pediatr Cardiol (2011) 32:724–736 731

123

of cases adds significant information to that obtained by

2D imaging. Furthermore, 3D ultrasonography can

detect extracardiac defects in the majority of patients

and can suggest the presence of malformation syn-

dromes. Image acquisition is rapid, minimally prolong-

ing the exam. Processing and analysis of the images

require more time and expertise, but the yield is pro-

portionally greater.

Although a number of reports [2, 18, 21, 22, 25, 26, 29]

have indicated advantages for 3D fetal imaging over 2D

imaging similar to those described in this report, a clear

role for 3D fetal imaging has yet to be established. Our

experience gives us the impression that 3D imaging is one

step closer to a direct physical examination of the fetus.

Just as the physical exam cannot always diagnose the

specific syndrome and must be confirmed by other means

such as karyotyping, so too, 3D ultrasonography requires

confirmation. In fact, a syndrome was suspected in eight of

nine fetuses, and a specific diagnosis was suggested in two

fetuses.

Examples of Syndromes Detected by 3D

Ultrasonography

Example 1 (Case 23 in Table 6)

A 3D exam performed at 23 gestational weeks because of

suspected CHD (and no other anomalous finding on 2D

examination) showed a hypoplastic nasal bone, upward

slant to the eyes, epicanthal folds of the eyelids, and an

anomalous oral fissure (tent-shaped lips), suggesting tri-

somy 21, which was confirmed at birth (Fig. 4).

Example 2 (Case 17 in Table 6)

A 3D examination of a 20-week-old fetus for high nuchal

translucency and single umbilical artery showed brachy-

cephaly, micrognathia, prominent occiput, diastatic sagittal

suture, and cleft lip. These findings suggested Edwards

syndrome (Fig. 5), which was confirmed by amniocentesis

and fetopsy.

Example 3 (Case 15 in Table 6)

The 3D images of a skeletal malformation in a 19-week-old

fetus with the 2D diagnosis of suspected CHD, microgna-

thia, and agenesia demonstrated micrognathia and an

anomalous left arm (Fig. 6a, b). A lateral view of the left

arm showed absence of the ulna (Fig. 6c, d).

Example 4 (Case 8 in Table 6)

The 3D examination of skull malformation in a 23-week-old

fetus detected micrognathia, diastatic sagittal suture, and

low-set ears (Fig. 7). Cornelia de Lange syndrome was

diagnosed postnatally, characterized by microcephaly,

micrognathia, small feet and hands, low-set ears, sagittal and

metopic diastatic suture, synophrys, and hypertrichosis.

The only fetus in which a syndrome was not detected

presented at 37 gestational weeks when adequate images of

the head and face could not be obtained. The most diag-

nostic images for general fetal morphology were obtained

in younger fetuses, generally before 24 gestational weeks

when termination of pregnancy still was an option.

In contrast to fetal morphology imaging, the cardiac

diagnostic definition was better for older fetuses, after 24

gestational weeks. This temporal limitation somehow

decreases the impact of 3D ultrasonography on the defi-

nition of the fetal cardiac anatomy. Three-dimensional

cardiac imaging was most useful for atrioventricular valves

and outflows.

Fig. 3 a Cleft lip well highlighted with three-dimensional ultraso-

nography (3DUS). b Postnatal comparison

732 Pediatr Cardiol (2011) 32:724–736

123

Examples of important new findings from the 3D exam

not present on the 2D ultrasonogram include a patent but

hypoplastic tricuspid valve in a fetus with a diagnosis of

tricuspid atresia by 2D ultrasonography (case 26 in

Table 4, Fig. 8, video 1 in supplementary material), dex-

trocardia in a fetus at 16 weeks of gestational age not

detected by 2D (case 20 in Table 4, Fig. 9, video 2 in

supplementary material), and aortic valve atresia in a

patient with aorto left ventricular tunnel (case 1 in Table 4,

Fig. 10, video 3 in supplementary material).

Adequate images of the aortic arch and the pulmonary

veins were rarely obtained, in part due to the lack of

dedicate acquisitions. A false image of a ventricular septal

defect (Fig. 11) occurred for one fetus, indicating a limi-

tation of this technique.

Despite the relative low surplus value of 3D versus 2D

ultrasonography with regard to the cardiac anatomy, we

think that what parents want to know above all is whether

their child will live a normal life. Therefore, the capability

of 3D imaging to detect features suggestive of genetic and

chromosomal syndromes is not of secondary importance.

Considering the high rate of such syndromes in fetuses

affected by CHD, we believe that a thorough 3D study

would be advisable for all the fetuses with a diagnosis of

CHD.

Even if the new generation of real-time 3D echocardi-

ography equipment allows a quicker reconstruction of the

image, we believe that 2D study will remain the first-line

approach for pediatric cardiologists studying the fetal heart,

whereas 3D ultrasonography should be reserved for all

cases in which a cardiac anomaly is detected.

In conclusion, our findings indicate that 3D fetal ultra-

sonography is a valuable tool for comprehensive evaluation

of the fetus. Besides providing information on the heart

anatomy using the three planes of space not obtainable with

2D echocardiography, it allows important features of the

somatic anatomy to be depicted in most cases. It thus

identifies those fetuses in whom a syndromic setting is

more likely present, allowing parents to receive more

Fig. 4 Down syndrome. b A three-dimensional (3D) exam of a

23-week-old fetus with suspected congenital heart defect (CHD)

showing hypoplastic nasal bone, upward slant to the eyes, epicanthal

folds of the eyelids, and an anomalous oral fissure (tent-shaped lips).

These findings suggest trisomy 21, which was confirmed at birth.

Note the normal nasal bone (a) compared with hypoplastic nasal bone

(b), and the postnatal view of the same patient (c)

Fig. 5 Edwards syndrome. A three-dimensional (3D) exam per-

formed at 20 gestational weeks (GW) because of high nuchal

translucency and a single umbilical artery shows brachicephaly,

micrognathia, prominent occiput, diastatic sagittal suture, and cleft

lip, suggesting Edwards syndrome, which was confirmed by amnio-

centesis and fetopsy

Pediatr Cardiol (2011) 32:724–736 733

123

Fig. 6 A 19-week-old fetus

with a two-dimensional (2D)

diagnosis of suspected

congenital heart defect (CHD),

micrognathia, and agenesia of

the ulna. The 3D images

demonstrate micrognathia

(a) and anomalous left arm (b).

The lateral view of the left arm

shows absence of the ulna (c, d)

Fig. 7 Cornelia de Lange

syndrome. a A three-

dimensional (3D) exam

performed at 23 gestational

weeks (GW) showing

micrognathia, diastatic sagittal

suture, and low-set ears.

Cornelia de Lange syndrome

was diagnosed postnatally,

characterized by microcephaly,

micrognathia, small feet and

hands, low-set ears, sagittal and

metopic diastatic suture,

synophrys, and hypertrichosis.

b Postnatal comparison

Fig. 8 Tricuspid hypoplasia in

a fetus (32 weeks gestational

age) affected by a univentricular

heart. The two-dimensional

(2D) image suggests tricuspid

atresia, whereas the 3D study at

the valvular plane demonstrated

that this valve was hypoplastic

but patent (refer video 1)

734 Pediatr Cardiol (2011) 32:724–736

123

comprehensive prenatal counseling, including the manda-

tory necessity of karyotyping in these specific cases. It is

useful to remember that the malformative syndrome rate is

very high for patients affected by CHD, accounting for

almost 25% of live-borns, and 70% of spontaneous abor-

tions and stillborns.

Fig. 9 Dextrocardia in a

16-week-old fetus clearly

showing the head, chest, and

heart (refer video 2)

Fig. 10 Valvular plane in a

24-week-old fetus affected by

aorto left ventricular tunnel,

ventricular septal defect, and

aortic atresia. a, b, c,

d Pulmonary valve (PA)

opening and closing according

to the cardiac cycle. The tunnel

is recognizable as a small

opening next to the aortic valve,

which is seen always closed and

firm. A clear view of this detail

was obtained only

retrospectively during a blind

review of the pediatric

cardiologist expert on three-

dimensional (3D) echo. The

atresia of the aortic valve was

not recognized on 2D

echocardiography but

discovered at the operating

table, with important clinical

implication. The understanding

of this defect would have

changed the surgical approach,

consequently improving the

outcome (refer video 3)

Fig. 11 False image of

ventricular septal defect (VSD).

a Labeled view of a false image

of apical VSD, referring to a

fetus of 21 gestational weeks

(GW) affected by tetralogy of

Fallot. Two apparent VSDs

seem well detectable, but only

one really exists. b Same image,

not labeled

Pediatr Cardiol (2011) 32:724–736 735

123

To our knowledge, this is the first study to outline the

role of 3D ultrasonography in a pediatric cardiology fetal

unit. With its specific limitations kept in mind, 3D ultra-

sonography should became a routine tool in the arma-

mentarium of the pediatric cardiologist, together with other

methods for fetal diagnosis.

Acknowledgment The authors gratefully acknowledge professor

Stephen P. Sanders for his valuable assistance in reviewing the

manuscript.

References

1. Baba K, Satoh K, Sakamoto S, Okai T, Ishii S (1989) Develop-

ment of an ultrasonic system for three-dimensional reconstruction

of the fetus. J Perinat Med 17:19–24

2. Bega G, Kuhlman K, Lev-Toaff A, Kurtz A, Wapner R (2001)

Application of three-dimensional ultraultrasonography in the

evaluation of the fetal heart. J Ultrasound Med 20:307–313

3. Bega G, Lev-Toaff A, Kuhlman K, Kurtz A, Goldberg B, Wapner

R (2001) Three-dimensional ultrasonographic imaging in

obstetrics: present and future applications. J Ultrasound Med 20:

391–408

4. Benacerraf BR (2002) Three-dimensional fetal ultrasonography:

use and misuse. J Ultrasound Med 21:1063–1067

5. Chaoui R, Kalache KD, Hartung J (2001) Application of three-

dimensional power Doppler ultrasound in prenatal diagnosis.

Ultrasound Obstet Gynecol 17:22–29

6. De Luca I (1999) Ecocardiografia fetale. In: Nicolosi GL (ed)

Trattato di ecocardiografia clinica 1999, vol 3. Piccin Publ,

Padova, pp 1983–2033

7. Deng J, Birkett AG, Kalache KD, Hanson MA, Peebles DM,

Linney AD et al (2001) Conversion of umbilical arterial Doppler

waveforms to cardiac cycle triggering signals: a preparatory study

for online motion-gated three-dimensional fetal echocardiogra-

phy. Ultrasound Med Biol 27:51–59

8. Deng J, Sullivan ID, Yates R, Vogel M, Mcdonald D, Linney AD

et al (2002) Real-time three-dimensional fetal echocardiography:

optimal imaging windows. Ultrasound Med Biol 28:1099–1105

9. DeVore GR, Falkensammer P, Sklansky MS, Platt LD (2003)

Spatiotemporal image correlation (STIC): a new technology for

evaluation of the fetal heart. Ultrasound Obstet Gynecol 22:380–

387

10. DeVore GR (2005) Three-dimensional and four-dimensional fetal

echocardiography: a new frontier. Curr Opin Pediatr 17:592–604

11. Espinoza J, Goncalves LF, Lee W, Mazor M, Romero R (2005) A

novel method to improve prenatal diagnosis of abnormal systemic

venous connections using three- and four-dimensional ultraultr-

asonography and ‘‘inversion mode’’. Ultrasound Obstet Gynecol

25:428–434

12. Ewigman BG, Crane JP, Frigoletto FD, LeFevre ML, Bain RP,

McNellis D (1993) Effect of prenatal ultrasound screening on

perinatal outcome. RADIUS study group. N Engl J Med 329:

821–827

13. Goncalves LF, Espinoza J, Kusanovic JP, Lee W, Nien JK,

Santolaya-Forgas J et al (2006) Applications of 2D matrix array

for 3D and 4D examination of the fetus: a pictorial essay.

J Ultrasound Med 25:745–755

14. Goncalves LF, Espinoza J, Lee W, Mazor M, Romero R (2004)

Three- and four-dimensional reconstruction of the aortic and

ductal arches using inversion mode: a new rendering algorithm

for visualization of fluid-filled anatomical structures. Ultrasound

Obstet Gynecol 24:696–698

15. Goncalves LF, Romero R, Espinoza J, Lee W, Treadwell M,

Chintala K et al (2004) Four-dimensional ultrasonography of the

fetal heart using color Doppler spatiotemporal image correlation.

J Ultrasound Med 23:473–481

16. Herberg U, Goldberg H, Breuer J (2003) Dynamic free-hand

three-dimensional fetal echocardiography gated by cardiotocog-

raphy. Ultrasound Obstet Gynecol 22:493–502

17. Houck RC, Cooke JE, Gill EA (2006) Live 3D echocardiography:

a replacement for traditional 2D echocardiography? AJR Am J

Roentgenol 187:1092–1106

18. Levental M, Pretorius DH, Sklansky MS, Budorick NE, Nelson

TR, Lou K (1998) Three-dimensional ultrasonography of normal

fetal heart: comparison with two-dimensional imaging. J Ultra-

sound Med 17:341–348

19. Levi S (2002) Ultrasound in prenatal diagnosis: polemics around

routine ultrasound screening for second trimester fetal malfor-

mations. Prenat Diagn 22:285–295

20. Linney AD, Deng J (1999) Three-dimensional morphometry in

ultrasound. Proc Inst Mech Eng H 213:235–245

21. Meyer-Wittkopf M, Cook A, McLennan A, Summers P, Sharland

GK, Maxwell DJ (1996) Evaluation of three-dimensional ultra-

sonography and magnetic resonance imaging in assessment of

congenital heart anomalies in fetal cardiac specimens. Ultrasound

Obstet Gynecol 8:303–308

22. Meyer-Wittkopf M, Rappe N, Sierra F, Barth H, Schmidt S

(2000) Three-dimensional (3D) ultrasonography for obtaining the

four and five-chamber view: comparison with cross-sectional

(2D) fetal sonographic screening. Ultrasound Obstet Gynecol

15:397–402

23. Nelson TR, Pretorius DH, Sklansky M, Hagen-Ansert S (1996)

Three-dimensional echocardiographic evaluation of fetal heart

anatomy and function: acquisition, analysis, and display. J Ultra-

sound Med 15:1–9

24. Scharf A, Geka F, Steinborn A, Frey H, Schlemmer A, Sohn C

(2000) 3D real-time imaging of the fetal heart. Fetal Diagn Ther

15:267–274

25. Sklansky MS, Nelson TR, Pretorius DH (1998) Three-dimen-

sional fetal echocardiography: gated versus non gated techniques.

J Ultrasound Med 17:451–457

26. Sklansky MS, Nelson TR, Pretorius DH (1997) Usefulness of

gated three-dimensional fetal echocardiography to reconstruct

and display structures not visualized with two-dimensional

imaging. Am J Cardiol 80:665–668

27. Tutschek B, David JS (2007) Three-dimensional echocardiogra-

phy for studies of the fetal heart: present status and future per-

spectives. Cardiol Clin 25:341–355

28. Yagel S, Cohen SM, Shapiro I, Valsky DV (2007) 3D and 4D

ultrasound in fetal cardiac scanning: a new look at the fetal heart.

Ultrasound Obstet Gynecol 29:81–95

29. Zosmer N, Jurkovic D, Jauniaux E, Gruboeck K, Lees C,

Campbell S (1996) Selection and identification of standard car-

diac views from three-dimensional volume scans of the fetal

thorax. J Ultrasound Med 15:25–32

736 Pediatr Cardiol (2011) 32:724–736

123