Applications of Advanced Ultrasound Technology in Obstetrics

diagnostics

Review Applications of Advanced Ultrasound Technology in Obstetrics

Kwok-Yin Leung

Obstetrics and Gynaecology, Gleneagles Hong Kong, Hong Kong, China; [email protected]

Abstract: Over the years, there have been several improvements in ultrasound technologies including high-resolution ultrasonography, linear transducer, radiant flow, three-/four-dimensional (3D/4D) ultrasound, speckle tracking of the fetal heart, and artificial intelligence. The aims of this review are to evaluate the use of these advanced technologies in obstetrics in the midst of new guidelines on and new techniques of obstetric ultrasonography. In particular, whether these technologies can improve the diagnostic capability, functional analysis, workflow, and ergonomics of obstetric ultrasound examinations will be discussed.

Keywords: obstetrics; ultrasound; 3D; 4D; Doppler; artiﬁcial intelligence

1. Introduction Ultrasound is widely used in obstetric practice to detect fetal abnormalities with a view to provide prenatal opportunities for further investigations including genetic testing and discussion of management options. In 2010, International Societies of Ultrasound in Obstetrics and Gynecology (ISUOG) published the practice guidelines on the minimal and optional requirements for a routine mid-trimester ultrasound scan [1]. Recently, The Citation: Leung, K.-Y. Applications American Institute of Ultrasound in Medicine (AIUM) suggests a detailed diagnostic of Advanced Ultrasound Technology second/third trimester scan for high-risk pregnancies [2], and fetal echocardiography for in Obstetrics. Diagnostics 2021, 11, at-risk pregnancies [3]. ISUOG has published recent guidelines on indications and practice 1217. https://doi.org/10.3390/ of targeted neurosonography [4,5]. Although the introduction of prenatal cell-free DNA- diagnostics11071217 based screening for Down syndrome has changed the role of the first trimester scan, the latter should still be offered to women [6]. Around 50% of major structural abnormalities Academic Editor: Ritsuko can be detected in the first trimester [7]. In addition, a recent study showed that a routine Kimata Pooh scan at around 36 weeks’ gestation can detect around 0.5% of previously undetected fetal abnormalities, as well as fetal growth restriction (FGR) [8]. Received: 7 June 2021 The detection rate of fetal abnormalities varies, depending on anatomy survey pro- Accepted: 3 July 2021 tocol, ultrasound equipment and setting, among other factors [9]. A high-resolution Published: 6 July 2021 ultrasound can facilitate a detailed diagnostic scan and a first-trimester scan and allow the detection of a small or subtle abnormality [10–12]. Although a detailed diagnostic scan is Publisher’s Note: MDPI stays neutral not required for all pregnant women, the indications include family history of congenital with regard to jurisdictional claims in malformation, maternal age 35 or above, gestational diabetes mellitus, artificial reproduc- published maps and institutional affil- tion technology, body mass index >= 30, teratogen, fetal nuchal translucency >= 3mm, and iations. many other conditions [2]. In the midst of such increasing standards of obstetric ultrasound examination, there is a demand on improving the diagnostic capability, functional analysis, workflow, and ergonomics. Over the years, there have been several improvements in ultrasound technologies including high-resolution ultrasonography, linear transducer, Copyright: © 2021 by the author. radiant flow, three/four-dimensional (3D/4D) ultrasound, speckle tracking of the fetal Licensee MDPI, Basel, Switzerland. heart, and artificial intelligence. The aim of this review is to evaluate the use of these This article is an open access article advanced technologies in obstetrics. distributed under the terms and conditions of the Creative Commons 2. High-Resolution Ultrasonography Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ High-resolution ultrasonography includes the use of a high-frequency transducer, 4.0/). and the means of enhancing image and signal processing including harmonic imaging

Diagnostics 2021, 11, 1217. https://doi.org/10.3390/diagnostics11071217 https://www.mdpi.com/journal/diagnostics Diagnostics 2021, 11, x FOR PEER REVIEW 2 of 18

2. High-Resolution Ultrasonography High-resolution ultrasonography includes the use of a high-frequency transducer, Diagnostics 2021, 11, 1217 and the means of enhancing image and signal processing including2 of harmonic 18 imaging (HI), spatial compound imaging (SCI), and speckle reduction imaging (SRI). Compared to a transducer with the low-frequency range (2 to 5 MHz), a transducer with the high-fre- (HI), spatialquency compound range imaging(5 to 9 MHz) (SCI), can and allow speckle for reduction improved imaging resolution (SRI). though Compared with limited tissue to a transducerpenetration. with the HI, low-frequency utilizing the rangephysics (2 toof 5non- MHz),linear a transducer propagation with of the ultrasound high- through the frequency range (5 to 9 MHz) can allow for improved resolution though with limited tissue body tissues, can produce high-resolution images with few artifacts. SCI, combining mul- penetration. HI, utilizing the physics of non-linear propagation of ultrasound through the bodytiple tissues, lines can of produce sight to high-resolution form a single images composite with fewimage artifacts. at real-time SCI, combining frame rates, can reduce multiple linesangle-dependent of sight to form artifacts. a single composite The use of image SRI atcan real-time reduce frame speckles rates, or can disturbances reduce that result angle-dependentfrom the artifacts. echo, which The use is projected of SRI can reducefrom an speckles ultrasound or disturbances transducer. that result from the echo, which is projected from an ultrasound transducer. 2.1. Fetal Echocardiography and Targeted Neurosonography 2.1. Fetal Echocardiography and Targeted Neurosonography ISUOG recommends the use of the highest possible transducer frequency to perform ISUOG recommends the use of the highest possible transducer frequency to perform fetal echocardiographyfetal echocardiography with a view with to improve a view the to likelihood improve of the detecting likelihood subtle of heart detecting de- subtle heart fects, albeitdefects, at the expensealbeit at of the reduced expense acoustic of reduced penetration acoustic [10] (Figure penetration1a–d and [10]Video (Figure S1). 1a–d and Video The use ofS1). HI The can improveuse of HI the can quality improve of ultrasound the quality images, of ultrasound especially when images, the maternalespecially when the ma- abdominalternal wall abdominal is thick during wall the is third thick trimester during ofthe pregnancy third trimester [11,13]. of pregnancy [11,13].

(a) (b)

Figure 1. High-resolutionFigure 1. ultrasonographyHigh-resolution ultrasonographyof the fetal heart of at the 20 fetalweeks’ heart gestation at 20 weeks’ showing gestation (a) a four-chamber showing (a) a view showing right atrium (RA),four-chamber left atrium view (LA), showing righ rightt ventricle atrium (RA),(RV), left and atrium left (LA),ventricle right (LV), ventricle (b) (RV), five-chamber and left ventricle view showing as- cending aorta (AAo)(LV), arising (b) ﬁve-chamber from the left view ventricle, showing ascendingthe right aortaand left (AAo) superior arising pulmonary from the left veins ventricle, (RSPV, the right LSPV) enter the left atrium (LA), andand descending left superior aorta pulmonary (DAo) veins behind (RSPV, the LSPV)LA (c) enter Three-vessel the left atrium view (LA),showing and descendingthe PA dividing aorta into the left (LPA) and right (RPA)(DAo) PA, behind AAo, the and LA the (c) Three-vessel superior vena view cava showing (SVC), the (d PA) three-vessel dividing into and the trachea left (LPA) view and showing right PA with the ductal branch (DA)(RPA) joining PA, AAo, the and DAo, the AAo, superior SVC, vena and cava trachea (SVC), (T); (d) Thymus three-vessel is anteri and tracheaor to the view three showing vessels. PA with the ductal branch (DA) joining the DAo, AAo, SVC, and trachea (T); Thymus is anterior to the three vessels. For a targeted neurosonographic examination, ISUOG recommends the use of high- resolution transvaginal transducers whenever possible [5]. An alternative is to use high-

Diagnostics 2021, 11, 1217 3 of 18

Diagnostics 2021, 11, x FOR PEER REVIEW 3 of 18

For a targeted neurosonographic examination, ISUOG recommends the use of high- resolution transvaginal transducers whenever possible [5]. An alternative is to use high- resolution transabdominal transducers with high frequency reaching 8–9 MHzMHz [[5].5]. TheThe anatomy ofof thethe fetalfetal brain brain is is examined examined in in details details on on a continuum a continuum of transverse, of transverse, sagittal sagittal and andcoronal coronal planes planes (Figure (Figure2a–d, 2a–d, Video Video S2a,b). S2a,b).

(a) (b)

Figure 2. High-resolution ultrasonography of the fetal brain at 20 weeks’ gestation: transverse views showing (a) posterior horn of the lateral ventricle (Vp), (b) cavum septi pellucidi (C.S.P.), ((cc)) cerebellumcerebellum (Cereb),(Cereb), CisternaCisterna magnamagna (CM),(CM), nuchalnuchal fold (NF),(NF), andand sagittal sagittal view view showing showing (d ()d corpus) corpus callosum callosum (CC), (CC), thalamus thalamus (TH), (TH), brain brain stem stem (BS), (BS), and cerebellarand cerebellar vermis vermis (CV). (CV). 2.2. Face and Neck 2.2. Face and Neck While the prenatal detection of cleft lip is high, the detection rate of subtle abnor- malitiesWhile of facethe prenatal such as low-setdetection or of posteriorly cleft lip is rotatinghigh, the ear, detection triangular rate face,of subtle down-slanting abnormal- itiespalpebral of face fissures, such as or low-set a long or and posteriorly marked philtrum rotating remainsear, triangular low [14 face,,15]. down-slanting These subtle abnor- pal- pebralmalities fissures, may be or features a long of and rare marked but severe philtrum genetic remains disorders low such [14,15]. as 5p These deletion subtle syndrome abnor- malitiesor RASopathy, may be whichfeatures require of rare chromosome but severe genetic microarray disorders analysis such as or 5p targeted deletion sequencing syndrome orfor RASopathy, RASopathy which genes. require As such, chromosome it is important microarray to perform analysis a detailed or targeted ultrasound sequencing scan for to RASopathyevaluate fetal genes. face As in fetusessuch, it especiallyis important if theyto perform have large a detailed NT, heart ultrasound defects, scan or unusual to eval- uatefindings fetal [face14,15 in]. fetuses High-resolution especially ultrasonographyif they have large allowsNT, heart the defects, clear visualization or unusual findings of facial [14,15].profile, lens,High-resolution nostrils, lips, ultrasonography maxilla, and ears allows (Figure the3 a–d,clear Videovisualization S3a,b). Recently,of facial profile, a new lens,sonographic nostrils,sign, lips, themaxilla, ‘superimposed and ears (Figure line’ sign, 3a–d, is suggested Video S3a,b). forevaluation Recently, a of new the sec-so- nographicondary palate sign, by the assessment ‘superimposed of the line’ vomeromaxillary sign, is suggested junction for evaluation in the midsagittal of the secondary view of palatethe palate by assessment [16] (Figure of3a). the vomeromaxillary junction in the midsagittal view of the palate [16] (Figure 3a).

Diagnostics 2021, 11, x FOR PEER REVIEW 4 of 18 DiagnosticsDiagnostics2021 2021, 11, 11, 1217, x FOR PEER REVIEW 4 ofof 1818

(a) (a) (b) (b)

(c) (c) (d) (d) Figure 3. High-resolutionFigure 3. High-resolution ultrasonographyFigure 3. High-resolution ultrasonography of the fetal ultrasonography face of theat 20 fetal weeks’ face of the gestation:at 20 fetal weeks’ face (a) at gestation:mid-sagittal 20 weeks’ (a gestation:) viewmid-sagittal showing (a)mid-sagittal view facial showing facial profile, (b) coronalprofile, view (b) coronalshowingview showingview both showing lens facial (Le), proﬁle,both (c) coronallens (b )(Le), coronal view (c) coronalshowing view showing view the uppershowing both lip lens the(UL) (Le), upper and (c) twolip coronal (UL) nostrils viewand (No), two showing nostrilsand the (No), and (d) sagittal view(d) showingsagittal view theupper ear.showing lip (UL) the and ear. two nostrils (No), and (d) sagittal view showing the ear.

LarynxLarynx and itsLarynx its movement movement and its can movement can be be assessed assessed can bybe prenatal byassessed prenatal ultrasoundby prenatal ultrasound (Figure ultrasound (Figure 4 and 4(Figure Videoand 4 and Video VideoS4). In S4 at-risk). InS4). at-risk fetuses In at-risk fetuses such fetuses as such those assuch with those as laryngeal those with laryngealwith atresia laryngeal atresia[17] andatresia [17 congenital] and [17] congenital and diaphrag- congenital di- diaphrag- aphragmaticmatic hernia,matic hernia,prenatal hernia, prenatal ultrasound prenatal ultrasound allowsultrasound systematic allows allows systematic examination systematic examination examination of the larynx, of theof including the larynx, larynx, including includingvocal cords vocal tovocal detect cords cords lary to detecttongeal detect laryngealanomalies laryngeal anomalies [17,18]. anomalies [17 [17,18].,18].

Figure 4. High-resolution ultrasonography of the fetal neck at 21 weeks’ gestation: sagittal view FigureFigure 4.4. High-resolutionHigh-resolution ultrasonographyultrasonography ofof thethe fetalfetal neckneck atat 2121 weeks’ weeks’ gestation: gestation: sagittalsagittal viewview showing larynxshowing (Lar) and larynx trachea (Lar) (T). and trachea (T). showing larynx (Lar) and trachea (T).

Diagnostics 2021, 11, 1217 5 of 18 Diagnostics 2021, 11, x FOR PEER REVIEW 5 of 18

2.3. Early Pregnancy Scan Transvaginal ultrasonography2.3. Early Pregnancy is Scan essential in the assessment of pregnancy of unknown location, which can beTransvaginal due to early ultrasonography pregnancy, miscarriage, is essential orin theectopic assessment pregnancy. of pregnancy It is of un- important to avoidknown making location, a false-positive which can diagnosisbe due to early of miscarriage pregnancy, miscarriage, by using transvaginal or ectopic pregnancy. sonography, carefulIt measurementis important to ofavoid mean making sac diameter a false-positive and crown diagnosis rump of length, miscarriage and using by using trans- safe cut-off valuesvaginal of these sonography, measurements careful inmeasurement defining miscarriage of mean sac diameter [19]. A recentand crown study rump length, and using safe cut-off values of these measurements in defining miscarriage [19]. A recent showed that amniotic sac sign (the presence of amniotic sac without a live embryo) is a study showed that amniotic sac sign (the presence of amniotic sac without a live embryo) reliable marker of miscarriage [20]. While the presence of an extrauterine gestational sac is a reliable marker of miscarriage [20]. While the presence of an extrauterine gestational with yolk sac and/orsac embryowith yolk with sac and/or or without embryo cardiac with or activity without is indicativecardiac activity of ectopic is indicative preg- of ectopic nancy, the presencepregnancy, of an inhomogeneous the presence of adnexal an inhomogeneous mass (‘blob’ adnexal sign) or mass extrauterine (‘blob’ sign) sac-like or extrauterine structure (‘bagel’ sign)sac-like is very structure suggestive (‘bagel’ of sign) a tubal is very ectopic suggestive pregnancy of a [tubal21]. Inectopic women pregnancy with [21]. In prior Caesarean section,women ultrasoundwith prior Caesarean features ofsection, Caesarean ultrasound scar pregnancyfeatures of Caesarean including scar low pregnancy implantation of theincluding gestational low sacimplantation within or of in the close gestational proximity sac within to a Caesarean or in close scar proximity as well to a Caesar- as classical signs ofean placenta scar as well accreta as classical spectrum signs disorders of placenta should accreta be spectrum looked outdisorders for [22 should,23]. be looked out for [22,23]. 2.4. First Trimester Scan 2.4. First Trimester Scan ISUOG and recently, AIUM published the practice guidelines on first-trimester fetal ultrasound scan [24,25].ISUOG High-resolution and recently, ultrasonographyAIUM published the allows practice the guidelines early assessment on first-trimester of fetal fetal anatomy [11]ultrasound (Figure5a–d, scan Video[24,25]. S5a,b)High-resolution and fetal ultrasonography malformations allows [ 12]. the Fetal early heart assessment of fetal anatomy [11] (Figure 5a–d, Video S5a,b) and fetal malformations [12]. Fetal heart can can be examined in the late first trimester [26], particularly with the use of color Doppler be examined in the late first trimester [26], particularly with the use of color Doppler (Video S5c,d). ISUOG recommends using high-frequency (6–12 MHz) transvaginal ul- (Video S5c,d). ISUOG recommends using high-frequency (6–12 MHz) transvaginal ultra- trasound to examinesound fetal to examine brain, especially fetal brain, if especially the focus if isthe in focus the posterioris in the posterior fossa and fossa the and the ma- maternal abdominalternal wall abdominal is thick [ wall5]. is thick [5].

(a) (b)

Figure 5. High-resolution ultrasonography of the fetus at 13 weeks’ gestation: (a) mid-sagittal view showing head, neck, and facial profile, (b) coronal view showing both eyes and ears, (c) the hand with five fingers, and (d) foot. Diagnostics 2021, 11, 1217 6 of 18

2.5. Doppler Ultrasound Doppler ultrasound is widely used in obstetrics. ISUOG has made recommendations on how to perform Doppler ultrasonography of the fetoplacental circulation [27]. It is a challenge to detect late-onset feral growth restriction (FGR). Although third-trimester– cerebroplacental ratio (CPR = middle cerebral artery pulsatile index/umbilical artery pulsatile index) is an independent predictor of stillbirth and perinatal mortality [28], CPR with or without adjustment for estimated fetal weight centile showed a low prediction rate for adverse perinatal outcome [29]. According to a meta-analysis, abnormal uterine artery (UtA) Doppler in the third trimester is useful in predicting perinatal death in suspected small-for-gestational age fetuses [30]. A recent prospective study suggested that cerebral– placental–uterine ratio (CPUR = CPR divided by mean UtA pulsatile index) detected FGR better than CPR or UtA Doppler alone [31].

2.6. Labour Ward Ultrasound The use of intrapartum ultrasonography is increasing. It can be performed by using a portable machine equipped with a wide-sector and low-frequency (<4 MHz) transducer and batteries with a long life, and being quick to start up and recharge [32]. According to ISUOG practice guidelines [32], intrapartum ultrasound is indicated when there is slow progress or arrest of labor in the ﬁrst or second stage. Recent studies showed that single ultrasound assessment of the fetal head station on admission in active phase or repeated measurements during active phase can predict the duration of labor and operative delivery in nulliparous women [33,34]. When the second stage of labor is prolonged, ultrasound can be used to assess fetal head position and station before considering or performing instrumental vaginal delivery [32]. Such assessment has a potential to predict mode of operative delivery and pregnancy outcomes [35]. Compared to clinical vaginal examinations, ultrasound assessment of the fetal head station and position is objective and reproducible [32–34], but assessment of cervical dilatation is limited when the dilatation is ≥ 4 cm and the membranes are ruptured [36]. Allowing detection of changes in tissue elasticity, elastography is a complementary technique to B-mode imaging, and it includes two methods, namely, shear-wave and strain elastography. A recent meta-analysis showed that the performance of cervical elastography was better than cervical length in the prediction of preterm delivery [37]. For the prediction of outcome of induction of labor, models based on inner cervical shear wave elastography and cervical length were more accurate than models based on the Bishop score [38].

3. Linear Transducer With a high-frequency ultrasound, a linear transducer can produce high-resolution images of shallow structures and small parts. Unlike curved transducers, linear transducers produce a rectangular field of view with uniform beam density throughout all tissue levels and without divergence in deeper tissue. The use of a transabdominal linear transducer can enhance the examination of the spinal cord and conus medullaris in the midsagittal view of the spine [5]. Some abnormalities such as cataract [39] and laryngeal atresia [17] can be well demonstrated using a linear transducer. A linear transducer can be used to examine fetal structures in the first trimester (Figure6a–d). However, a linear transducer is not suitable for using if the structures of interest are deep or the maternal abdominal wall is thick. Although a linear transducer can allow the examination of the fetal cardiac anatomy at 11–13 weeks [40], it is the use of color flow mapping but not of a linear transducer that improves the examination [41]. Diagnostics 2021, 11, x FOR PEER REVIEW 7 of 18 Diagnostics 2021, 11 , 1217 7 of 18

(a) (b)

Figure 6. UltrasonographyFigure 6. Ultrasonography of a fetus at of13 a weeks’ fetus at gestation 13 weeks’ by gestation a transabdominal by a transabdominal high-frequency high-frequency linear transducer: (a) transverse viewlinear of transducer:fetal brain, ( ba)) transversecoronal view view of face of fetal showing brain, both (b) coronal orbits (OB), view ( ofc) facecoronal showing view of both abdomen orbits showing both kidneys(OB), (Ki) (onc) coronaleither side view of ofthe abdomen spine, and showing (d) the three-vessel both kidneys trachea (Ki) ontransverse either side view of with the color spine, Doppler and (d s)howing the pulmonary artery (PA), aorta (Ao), superior vena cava (SVC), and trachea (T). three-vessel trachea transverse view with color Doppler showing pulmonary artery (PA), aorta (Ao), superior vena cava (SVC), and trachea (T). 4. Radiant Flow 4. Radiant Flow Radiant flow shows the blood flow with a sense of depth by using a specific algo- Radiant flowrithm shows to convert the blood the flow index with of aerythrocyte sense of depth density by using in a acertain specific area algorithm into a height index to convert thewhich index is of then erythrocyte superimposed density on in the a initial certain coding area into of color, a height power index Doppler, which or high-defini- is then superimposedtion flow on[42]. the Other initial advantages coding of include color, power reducing Doppler, blood overflow or high-definition and indicating the ves- flow [42]. Othersel advantageswith sharp edges. include Special reducing display blood produced overflow by andsimilar indicating technologies the vessel include MicroFlow with sharp edges.Imaging Special (Philips), display MV-Flow, produced and by LumiFlow similar technologies (Samsung). include MicroFlow Imaging (Philips), MV-Flow,Radiant flow and is LumiFlow used to show (Samsung). fast blood flow in the fetal heart and brachycephalic Radiant flowarteries is used [42] (Videos to show S5d fast and blood S6a,b), flow as in well the fetalas slow-blood heart and flow brachycephalic in the neurovascular cir- arteries [42] (Videosculation S5d [43] and(Video S6a,b), S7). as well as slow-blood flow in the neurovascular circulation [43] (VideoThe S7).fetal umbilical–portal venous system is complex. High-definition flow imaging The fetal(HDFI) umbilical–portal has been used venous to systemassess the is complex. normal anatomy High-definition of this system flow imaging or umbilical–portal– (HDFI) has beensystemic used tovenous assess shunts. the normal Transverse anatomy and of sagitta this systeml planes or are umbilical–portal– used to examine the fetal um- systemic venousbilical–portal shunts. Transverse venous system and (Video sagittal S8a,b). planes In are a recent used case to examine report, the the authors fetal used HDFI umbilical–portaland venous radiant system flow imaging (Video S8a,b).to clearly In delineate a recent case the report,aberrant the course authors of the used ductus venosus HDFI and radiantreturning flow to imaging the coronary to clearly sinus delineate [44]. the aberrant course of the ductus venosus returning to the coronary sinus [44]. 5. 3D/4D Ultrasound 5. 3D/4D Ultrasound Over the years, new 2D modes (such as high-density power imaging), new 3D vol- Over the years, new 2D modes (such as high-density power imaging), new 3D volume ume acquisition (such as Corpus callosum mode or matrix probe), and new analysis (such acquisition (such as Corpus callosum mode or matrix probe), and new analysis (such as as semiautomated analysis) have been added in 3D/4D ultrasound examinations (Table semiautomated analysis) have been added in 3D/4D ultrasound examinations (Table1). 1). The use of 3D multiplanar/multislice analysis facilitates the assessment of normal and

Diagnostics 2021, 11, 1217 8 of 18

The use of 3D multiplanar/multislice analysis facilitates the assessment of normal and abnormal structures in standard planes. This can also facilitate the detection of subtle fetal defects [45]. The use of 3D rendered images can help counseling to the women when fetal malformations are found or reassure the at-risk women when normal fetal anatomy is found [45]. 3D/4D US is useful for the assessment of fetal brain, spine, face, heart, and other structures [45,46].

Table 1. Commonly used scanning mode, volume acquisition, and analysis for three-/four- dimensional (3D/4D) ultrasound examinations.

Mode Volume Acquisition 3D/4D Analysis Gray scale 3D: different modes Multiplanar Color ﬂow 4D Multislice Power doppler STIC 1 Rendered view: different modes High-density power imaging Matrix probe Cine loop B-ﬂow Semi-automatic analysis Volume measurement Power Doppler measurements 1 Spatiotemporal image correlation.

Examples in the assessment of fetal abnormalities a. Cleft lip and palate: use gray-scale mode, after a 3D volume acquisition, perform multiplanar/multi-slice analysis and rendering techniques to assess the integrity of palate. b. Short-limbed and short-rib dysplasia: use gray-scale mode, after a 3D volume acquisition with skeletal mode, perform rendering techniques with skeletal mode to examine the long bones and ribs. c. Agenesis of ductus venosus: use high-density power imaging, after a 3D volume acquisition, perform multi-slice analysis to assess the precordial venous system. d. Cardiac outﬂow tract abnormalities: use color ﬂow, after a STIC volume acquisition, perform multiplanar/multi-slice analysis in a cine-loop of cardiac cycle. e. Atrioventricular valve abnormalities: use matrix probe and gray-scale mode, real- time 4D cine-loop analysis to display the coronal view of atrioventricular valve.

5.1. 3D Neurosonography In targeted neurosonography, a systematic assessment of the fetal brain is required. Although this assessment can be performed by a 2D ultrasound examination, a perfect midsagittal view may not be achieved at all times, thus affecting a proper assessment. ISUOG recommends using 3D ultrasound examination that can provide images of enhanced resolution by displaying thicker ‘slices’ of the brain and thus increasing the signal-to- background noise ratio on all three planes. In addition, multiplanar imaging correlation allows the display of perfectly aligned views on the three orthogonal planes [5]. To avoid shallowing by adjacent skull bones, it is important to acquire a 3D volume in a mid- sagittal plane through the sagittal suture. If the focus is on the anterior complex, the volume will be obtained from the anterior fontanelle or the anterior part of the sagittal suture [5] (Figure7a). If the focus is on the posterior fossa and cerebellar vermis, the volume will be obtained from the posterior fontanelle or the posterior part of the sagittal suture with the ultrasound beam being almost perpendicular to the tentorium [47] (Figure7b). A transvaginal and transabdominal approach is used when the fetal presentation is vertex and breech, respectively. Then, the midlines structures including corpus callosum, brain stem, and cerebellar vermis can be examined by multiplanar and multi-slice analysis [43,48,49]. An accurate measurement of corpus callosum and cerebellar vermis can be achieved. DiagnosticsDiagnostics 2021 2021, 11, ,11 x ,FOR x FOR PEER PEER REVIEW REVIEW 9 of 918 of 18 Diagnostics 2021, 11, 1217 9 of 18

(a)( a) (b) (b) FigureFigureFigure 7.7. Three-dimensional7.Three-dimensional Three-dimensional ultrasonography ultrasonography of fe fetaloftal fe braintal brain brain at at 20 20at weeks’20 weeks’ weeks’ gestation: gestation: gestation: (a (a) ) multiplanar( multiplanara) multiplanar analysis analysis analysis after after after a avolume volume a volume acquisitionacquisitionacquisition withwith with corpuscorpus corpus callosum callosum callosum mode mode mode through through through the the anteriorthe anterior anterior part part ofpart of the theof sagittal thesagittal sagittal suture suture suture showing showing showing corpus corpus corpus callosum callosum callosum (CC), (CC), and (CC), and (b) multiplanar analysis after a volume acquisition through the posterior fontanelle showing corpus callosum (CC), (b)and multiplanar (b) multiplanar analysis analysis after a volume after a volume acquisition acquisition through through the posterior the posterior fontanelle fontanelle showing corpusshowing callosum corpus (CC),callosum cavum (CC), cavum septi pellucidi (C.S.P.), thalamus (TH), brainstem (BS), and cerebellar vermis (CV). septicavum pellucidi septi (C.S.P.),pellucidi thalamus (C.S.P.), (TH),thalamus brainstem (TH), brainstem (BS), and cerebellar (BS), and vermiscerebellar (CV). vermis (CV).

AfterAfter a a 3D 3Da 3D volume volume volume acquisition acquisition acquisition of of the of the the fetal fetal fetal sp spineine sp ineat at mid-sagittal at mid-sagittal mid-sagittal plane, plane, plane, a rendered a rendereda rendered view view view of the fetal spine can be well displayed with various modes (Figure 8a,b). In addition, the ofof the the fetal fetal spine spine can can be be well well displayeddisplayed withwith various modes (Figure (Figure 88a,b).a,b). In In addition, addition, the coronal planes at the level of the vertebral bodies and/or posterior arches can be recon- thecoronal coronal planes planes at atthe the level level of ofthe the vertebral vertebral bodies bodies and/or and/or posterior posterior arches arches can be can recon- be structed on multiplanar analysis [5]. reconstructedstructed on multiplanar on multiplanar analysis analysis [5]. [5].

(a) (b) (a) (b) Figure 8. Three-dimensional rendered views of fetal spine at 20 weeks’ gestation after a volume acquisition with skeletal mode:Figure (a) 8. usual Three-dimensional mode, and (Figureb) X-ray rendered 8. mode.Three-dimensional views of fetal spine rendered at 20 weeks’ views ofgestation fetal spine after at a 20volume weeks’ acquisition gestation with after skeletal a volume mode: (a) usual mode, andacquisition (b) X-ray withmode. skeletal mode: (a) usual mode, and (b) X-ray mode. 5.2. Spatiotemporal Image Correlation 5.2.5.2. SpatiotemporalSpatiotemporal Spatiotemporal Image ImageImage Correlation CorrelationCorrelation (STIC) allows an automatic acquisition of a single 3D volumeSpatiotemporalSpatiotemporal through slow Image Image sweep Correlation Correlation and subsequent (STIC) (STIC) allows allowsanalysis an an automatic in automatic a looped acquisition acquisition cine sequency of of a singlea ofsingle images3D3D volume volume in the through multiplanar/multi-slicethrough slow slow sweep sweep and and format subsequent subsequent and a rendered analysis analysis inview. in a loopeda Thislooped can cine cine produce sequency sequency im- of of agesimagesimages in ina standardized thein the multiplanar/multi-slice multiplanar/multi-slice plane while minimizi format format andng andthe a rendered operationa rendered view. dependency view. This canThis produce canof the produce ultra- images im- soundinages a standardized examination. in a standardized plane The recent while plane advances minimizing while minimizi in gr theay operationscaleng the and operation color dependency Doppler dependency of post the processing ultrasound of the ultra- examination.sound examination. The recent The advances recent inadvances gray scale in andgray color scale Doppler and color post Doppler processing post improves processing

Diagnostics 2021, 11, x FOR PEER REVIEW 10 of 18

Diagnostics 2021, 11, 1217 10 of 18

theimproves display the of ultrasound display of images. ultrasound Using images. color Doppler Using color with STICDoppler in the with glass-body STIC in mode the glass- canbody show mode normal can andshow abnormal normal anatomyand abnormal of the fetalanatomy heart andof the major fetal vessels heart [and46] (Figuremajor 9vessels, Video[46] (Figure S9). The 9, matrix Video probe S9). The allows matrix the rapidprobe acquisition allows the ofrapid an STIC acquisition volume, of thus an reducingSTIC volume, thethus motion reducing artifact the and motion facilitating artifact live and 4D facilitating display [46 live]. In 4D addition, display the [46]. use In of addition, the matrix the use probeof the allows matrix the probe simultaneous allows the examination simultaneous of two examination orthogonal of planes two orthogonal of the fetal planes heart in of the thefetal ‘biplane heart mode’.in the ‘biplane Additional mode’. use of Additional image recognition use of image software recognition can help review software cardiac can help structuresreview cardiac in the standardstructures planes in the [ 46standard]. The 3D plan renderedes [46]. images The 3D are rendered useful for images counseling are useful tofor parents. counseling In addition, to parents. STIC In addition, volume can STIC facilitate volume interdisciplinary can facilitate interdisciplinary consultation and consul- teleconsultationtation and teleconsultation [42]. [42].

FigureFigure 9. 9.Color Color Doppler Doppler with with spatiotemporal spatiotemporal image image correlation correlation in the in glass-bodythe glass-body mode mode showing showing multiplanarmultiplanar view view and and a rendereda rendered image image of aof normal a normal fetal fetal heart heart at 20 at weeks’ 20 weeks’ gestation. gestation.

5.3.5.3. 3D 3D Ultrasound Ultrasound Examination Examination of Face,of Face, Limbs, Limbs, and and Other Other Structures Structures WhileWhile 2D 2D ultrasound ultrasound is ais keya key tool tool for for the the detection detection of fetal of fetal anomalies, anomalies, there there are some are some anomaliesanomalies such such as as facial facial clefts, clefts, micrognathia, micrognathia, and and club club foot foot in which in which 3D ultrasound 3D ultrasound may may provideprovide additional additional information information or or help help counseling counseling when when such such anomaly anomaly is suspected is suspected [9]. [9]. Compared to 2D ultrasound alone, combined approach of 2D and 3D ultrasound with Compared to 2D ultrasound alone, combined approach of 2D and 3D ultrasound with multiplanar/multi-slice analysis can improve the detection or exclusion of cleft palate multiplanar/multi-slice analysis can improve the detection or exclusion of cleft palate in in fetuses with cleft lip [50] (Figure 10). Although 3D ultrasound is less sensitive for the fetuses with cleft lip [50] (Figure 10). Although 3D ultrasound is less sensitive for the de- detection of isolated cleft palate, a recent study showed that an accurate evaluation of palate requirestection of 3D isolated ultrasound cleft examination palate, a recent with study volume showed acquisition that an in aaccurate strictly axialevaluation transverse of palate viewrequires of the 3D palate ultrasound [16]. The examination use of 3D ultrasound with volume multiplanar acquisition analysis in a strictly and 3D axial rendering transverse viewview can of facilitatethe palate the [16]. display The ofuse mid-sagittal of 3D ultrasound plane of multiplanar the fetal face analysis and thus and improve 3D rendering the accuracyview can of facilitate measurements the display of the of mandible mid-sagittal and the plane detection of the offetal micrognathia face and thus [51]. improve the accuracyThree-dimensional of measurements (3D) of rendering the mandible technology and the with detection skeletal modeof micrognathia can display [51]. skull, vertebrae, ribs, long bones and ﬁngers [52] (Figure 11a,b and Video S10a,b). Prenatal assessment of the ribs and vertebral pattern can be performed by 3D ultrasound with skeletal mode (Figure8a,b), albeit it is not a routine practice. A review of 39 studies including 75,018 healthy subjects and 6130 subjects with structural or chromosomal anomalies or adverse outcome showed an association between cervical ribs and other structural anomalies including esophageal atresia and anorectal malformation [53]. Abnormalities such as craniosynostosis [26,54], and extra ribs can be shown.

Diagnostics 2021, 11, x FOR PEER REVIEW 11 of 18

Diagnostics 2021, 11, x FOR PEER REVIEW 11 of 18 Diagnostics 2021, 11, 1217 11 of 18

Figure 10. Three-dimensional ultrasound assessment of the fetal face at 13 weeks’ gestation showing multiplanar views of lip and palate (reference dot).

Three-dimensional (3D) rendering technology with skeletal mode can display skull, vertebrae, ribs, long bones and fingers [52] (Figure 11a,b and Video S10a,b). Prenatal assessment of the ribs and vertebral pattern can be performed by 3D ultrasound with skeletal mode (Figure 8a,b), albeit it is not a routine practice. A review of 39 studies including 75,018 healthy subjects and 6,130 subjects with structural or chromosomal anomalies or

adverse outcome showed an association between cervical ribs and other structural anom- FigureFigure 10. 10.Three-dimensional Three-dimensionalalies including ultrasound ultrasound esophageal assessment assessment atresia of ofand the the fetalanorectal fetal face face at atmalformation 13 13 weeks’ weeks’ gestation gestation [53]. showing Abnormalitiesshowing such as multiplanar views of lip and palate (reference dot). multiplanar viewscraniosynostosis of lip and palate (reference[26,54], and dot). extra ribs can be shown. Three-dimensional (3D) rendering technology with skeletal mode can display skull, vertebrae, ribs, long bones and fingers [52] (Figure 11a,b and Video S10a,b). Prenatal assessment of the ribs and vertebral pattern can be performed by 3D ultrasound with skeletal mode (Figure 8a,b), albeit it is not a routine practice. A review of 39 studies including 75,018 healthy subjects and 6,130 subjects with structural or chromosomal anomalies or adverse outcome showed an association between cervical ribs and other structural anomalies including esophageal atresia and anorectal malformation [53]. Abnormalities such as craniosynostosis [26,54], and extra ribs can be shown.

(a) (b)

Figure 11. Three-dimensionalFigure 11. Three-dimensional rendered images rendered of the images fetal skeleton of the at fetal 20–22 skeleton weeks’ at gestation 20–22 weeks’ showing: gestation (a) X-ray mode of the skull boneshowing: with frontal (a) X-ray suture mode (SS) of theand skull anterior bone fontanelle with frontal (AF), suture and (SS) (b) andHD anteriorskeletal fontanellemode of the (AF), skull, and bones of the upper and (lowerb) HD limbs. skeletal mode of the skull, bones of the upper and lower limbs.

It is difﬁcult toIt visualize is difficult esophagus to visualize on 2D esophagus ultrasound on examination. 2D ultrasound The examination. use of 3D ul- The use of 3D trasound with multiplanarultrasound with analysis multiplanar and Crystal analysis Vue rendering and Crystal may Vue make rendering the visualization may make the visualiza- possible [55]. Three-dimensionaltion possible [55]. (3D)Three-dimensional volumes are acquired (3D) volumes from a midsagittalare acquired section from a of midsagittal sec- the thorax(a) andtion upper of abdomenthe thorax with and theupper fetus abdomen lying in(b with supine) the position. fetus lying in supine position. While 2D ultrasoundWhile examination 2D ultrasound with examination gray scale and with color gray ﬂow scale is theand standard color flow for is the the standard for Figure 11. Three-dimensionalantenatal rendered diagnosis imagesthe antenatalof of the placenta fetal diagnosisskeleton accreta atspectrum of20–22 placenta weeks’ disorders accreta gestation [ 22spectrum showing:], 3D ultrasound disorders (a) X-ray withmode [22], power of3D ultrasound with the skull bone with frontalDoppler suture (SS) and and multiplanar anterior fontanelle analysis (AF), permits and an(b) accurateHD skeletal assessment mode of the of theskull, placenta-bladder bones of the upper and lower limbs. interface, and the degree of bladder invasion by the placenta [56]. Three-dimensional (3D) rendered images can be used for patient counseling [56]. It is difficult to visualize esophagus on 2D ultrasound examination. The use of 3D 5.4.ultrasound 3D Printing with multiplanar analysis and Crystal Vue rendering may make the visualiza- tionWith possible advances [55]. Three-dimensional in 3D ultrasound, (3D) thederived volumes ultrasound are acquired data from can abe midsagittal used for 3Dsec- printingtion of the of physicalthorax and models upper of abdomen whole fetuses with the [57 fetus] and lying the fetalin supine face [position.58]. A recent trial showedWhile that 2D the ultrasound use of 3D-printed examination fetal with facial gray models scale and resulted color inflow greater is the increases standard infor the antenatal diagnosis of placenta accreta spectrum disorders [22], 3D ultrasound with

Diagnostics 2021, 11, x FOR PEER REVIEW 12 of 18

power Doppler and multiplanar analysis permits an accurate assessment of the placenta- bladder interface, and the degree of bladder invasion by the placenta [56]. Three-dimensional (3D) rendered images can be used for patient counseling [56].

5.4. 3D Printing With advances in 3D ultrasound, the derived ultrasound data can be used for 3D printing of physical models of whole fetuses [57] and the fetal face [58]. A recent trial Diagnostics 2021, 11, 1217 showed that the use of 3D-printed fetal facial models resulted in greater increases12 of 18 in maternal–fetal attachment in the third trimester than the use of ultrasonography only [58]. Whether this can be translated into better pregnancy outcomes needs further studies. In maternal–fetal attachment in the third trimester than the use of ultrasonography only [58]. Whetheraddition, this a 3D-printed can be translated spina into bifida better model pregnancy can be outcomes beneficial needs for surgical further studies. rehearsal In prior to addition,a fetoscopic a 3D-printed repair [59]. spina bifida model can be beneficial for surgical rehearsal prior to a fetoscopicWith repair advances [59]. in STIC, the derived data can be used for 3D printing of the fetal heart, whichWith is advancesa fast-moving in STIC, structure the derived [60]. data In can a recent be used case for 3Dreport, printing the of authors the fetal found heart, that the which3D model is a fast-moving was useful structure in showing [60]. In the a recent complex case anatomy report, the of authors fetal transposition found that the 3Dof great ar- modelteries wasand useful in providing in showing prenatal the complex parental anatomy counseling of fetal [61]. transposition of great arteries and inPreviously, providing prenatal after acquisition parental counseling of a 3D/STIC [61]. volume dataset, a number of post-pro- Previously, after acquisition of a 3D/STIC volume dataset, a number of post-processing cessing steps are required to convert it from Cartesian.vol file through segmentation, re- steps are required to convert it from Cartesian.vol file through segmentation, refinement, andfinement, optimization and optimization to a STL (Standard to a TriangleSTL (Standard Language) Triangle file, the industryLanguage) standard file, the file industry typestandard for 3D file Printing type [for60]. 3D These Printing steps take [60]. a longThese time, steps and take whether a long the time, final producedand whether STL the final fileproduced is good enoughSTL file for is 3D good printing enough is not for certain 3D prin beforeting processing. is not certain With before recent advancesprocessing. With inrecent ultrasound advances technology, in ultrasound a 3D/STIC technology, volume dataset a 3D/STIC can be volume directly dataset exported can from be thedirectly ex- ultrasoundported from machine the ultrasound as an STL file machine that is readyas an forSTL viewing file that on is a ready personal for computer viewingusing on a personal commoncomputer software using ascommon well as forsoftware 3D printing as well (Figure as for 12 ).3D printing (Figure 12).

FigureFigure 12. 12.A A physical physical model model of three-dimensional of three-dimensional printing printing of the fetalof the face. fetal face.

6.6. FetalHQ FetalHQ FetalHQ, a novel heart and vascular analysis software, can allow assessment of the FetalHQ, a novel heart and vascular analysis software, can allow assessment of the fetal heart shape, size, and contractibility by using speckle tracking to analyze the motion of multiplefetal heart points shape, of the size, fetal and heart contractibility [62] (Figure 13). by The using global speckle sphericity tracking index (SI)to analyze is a simple the motion measurementof multiple points of cardiac of the contractility, fetal heart and [62] it is(Figure equal to13). (end-diastolic The global mid-basal–apicalsphericity index (SI) is a length)/transversesimple measurement length of [63 cardiac]. For 24-segment contractility, sphericity and it index, is equal SI is to computed (end-diastolic for each mid-basal– of theapical 24 end-diastolic length)/transverse transverse length segments, [63]. whichFor 24-s areegment distributed sphericity from the index, base toSI the is computed apex for ofeach each of ventricle, the 24 end-diastolic as well as the end-diastolictransverse segments, mid-basal–apical which length are distributed [62]. from the base to the Thisapex 24-segment of each ventricle, sphericity as indexwell as is athe comprehensive end-diastolic method mid-basal–apical to assess the length shape of[62]. ventricular chambers [62]. The SI for each segment was independent of gestational age and fetal biometry. The SI of the right ventricle was lower than that of the left ventricle for segments 1–18. This index can be used when discordance between the size of the atrial and/or ventricular cardiac chambers is found. Abnormal SI values are found in the fetuses with cardiac abnormalities such as coarctation of aorta, pulmonary stenosis, and fetal growth restriction [62]. Abnormal SI values are associated with increased risk of perinatal complications and childhood and/or adult cardiovascular disease [64].

Diagnostics 2021, 11, x FOR PEER REVIEW 13 of 18 Diagnostics 2021, 11, 1217 13 of 18

FigureFigure 13.13. FetalHQFetalHQ assessmentassessment ofof thethe fetalfetal heartheart shape,shape, size,size, andand contractibility contractibility byby using using speckle speckle tracking to analyze the motion of multiple points of the fetal heart at 20 weeks’ gestation. tracking to analyze the motion of multiple points of the fetal heart at 20 weeks’ gestation.

WhileThis 24-segment the initial results sphericity are promising, index is a a co recentmprehensive review of method 23 studies to showedassess the conflicting shape of resultsventricular concerning chambers the development[62]. The SI for of each strain segment and strain was rate independent during gestation of gestational [65]. Large age longitudinaland fetal biometry. cohort studiesThe SI of with the aright standard ventricle protocol was lower are needed than that to obtain of the reference left ventricle values for forsegments fetal cardiac 1–18. This deformation index can in be uncomplicated used when discordance pregnancies between [65]. Athe recent size of systematic the atrial reviewand/or alsoventricular showed cardiac heterogeneous chambers results is found. concerning Abnormal gestational SI values age are and found Doppler in the profiles. fetuses Largewith prospectivecardiac abnormalities longitudinal such cohort as studiescoarctation are requiredof aorta, to pulmonary assess the clinical stenosis, significance and fetal ofgrowth deformation restriction measurements [62]. Abnormal of the SI values fetal heart are associated in growth with restricted increased fetuses risk and of perinatal normal fetusescomplications [66]. and childhood and/or adult cardiovascular disease [64]. While the initial results are promising, a recent review of 23 studies showed conflict- 7.ing Artificial results Intelligenceconcerning the development of strain and strain rate during gestation [65]. LargeMachine longitudinal learning, cohort in particularstudies with deep a standard learning, protocol allows ultrasoundare needed imageto obtain recognition reference andvalues thus for facilitates fetal cardiac the automatic deformation identification in uncomplicated and measurement pregnancies of fetal[65]. biometryA recent system- [67]. It isatic a branch review of also artificial showed intelligence heterogeneous (AI). Inresu obstetriclts concerning ultrasonography, gestational the age automation and Doppler of measurementsprofiles. Large ofprospective fetal biometry longitudinal is a potentially cohort usefulstudies tool are torequired increase to the assess reliability the clinical and reproducibilitysignificance of ofdeformation measurements measurements as compared of to the manual fetal measurementsheart in growth [68 restricted]. In addition, fetuses it canand reduce normal scanning fetuses [66]. time [68] and work-related fatigue and musculoskeletal disorders [69]. With automatic image recognition technology applied on a frozen 2D ultrasound image,7. Artificial auto measurementIntelligence of fetal biometry including biparietal diameter (BPD), head circumferenceMachine (HC),learning, abdominal in particular circumference deep learning, (AC), allows and femur ultrasound length (FL)image becomes recognition feasible.and thus A study facilitates showed the aautomatic success rate identification of 91.43% andand 100%measurement for auto of measurement fetal biometry of HC[67]. andIt is BPD,a branch respectively of artificial [67 intelligence]. Although (AI). the In inaccuracy obstetric ultrasonography, for the plane acceptance the automation check for of headmeasurements parameters of was fetal 12.9% biometry [67], is such a potentially inaccuracy useful can be tool corrected to increase by fine-tuningthe reliability of theand caliperreproducibility placement of manually.measurements In another as compared study, manual to manual adjustment measurements of caliper [68]. position In addition, was notit can required reduce in scanning about two-thirds time [68] of and cases work-related for HC and fatigue FL measurements, and musculoskeletal but it was disorders required in[69]. more than 80% for the measurement of AC [68]. Auto measuring AC is more difficult than measuringWith automatic HC because image recognition of the low contrasttechnology between applied the on abdomen a frozen 2D and ultrasound surrounding im- tissues and the large variability in abdominal shape and appearance [68]. The accuracy of age, auto measurement of fetal biometry including biparietal diameter (BPD), head cir- the auto measurement of HC, AC, and FL was high, and it compared well with previously cumference (HC), abdominal circumference (AC), and femur length (FL) becomes feasi- published manual-to-manual agreement, but the auto measurements had a tendency to ble. A study showed a success rate of 91.43% and 100% for auto measurement of HC and underestimate biometry, which requires further improvements in the algorithm [68]. BPD, respectively [67]. Although the inaccuracy for the plane acceptance check for head With automatic image recognition technology applied on an acquired 3D ultrasound parameters was 12.9% [67], such inaccuracy can be corrected by fine-tuning of the caliper volume of the fetal head from the BPD plane, SonoCNS allows auto measurement of fetal placement manually. In another study, manual adjustment of caliper position was not re- biometry including BPD, HC, atrium of the posterior horn of the lateral ventricle (Vp), quired in about two-thirds of cases for HC and FL measurements, but it was required in transcerebellar diameter (TCD), and cisterna magnum (CM) [70] (Figure 14). A recent more than 80% for the measurement of AC [68]. Auto measuring AC is more difficult than study showed that this 3D automated technology reliably identified and measured BPD measuring HC because of the low contrast between the abdomen and surrounding tissues and the large variability in abdominal shape and appearance [68]. The accuracy of the auto

Diagnostics 2021, 11, x FOR PEER REVIEW 14 of 18

measurement of HC, AC, and FL was high, and it compared well with previously published manual-to-manual agreement, but the auto measurements had a tendency to underestimate biometry, which requires further improvements in the algorithm [68]. With automatic image recognition technology applied on an acquired 3D ultrasound volume of the fetal head from the BPD plane, SonoCNS allows auto measurement of fetal Diagnostics 2021, 11, 1217 biometry including BPD, HC, atrium of the posterior horn of the lateral ventricle14 (Vp), of 18 transcerebellar diameter (TCD), and cisterna magnum (CM) [70] (Figure 14). A recent study showed that this 3D automated technology reliably identified and measured BPD andand HC HC but but was was less less so so for for TCD, TCD, CM, CM, and and Vp Vp [ 70[70].]. Further Further optimization optimization of of this this automated automated technologytechnology is is required. required.

FigureFigure 14. 14. SonoCNS,SonoCNS, after volume acquisition of of the the feta fetall brain brain at at biparietal biparietal diameter diameter plane plane at at 21 weeks gestation, showing auto measurement of biparietal diameter (BPD), head circumference 21 weeks gestation, showing auto measurement of biparietal diameter (BPD), head circumference (HC), atrium of the posterior horn of the lateral ventricle (Vp), transcerebellar diameter (TCD), and (HC), atrium of the posterior horn of the lateral ventricle (Vp), transcerebellar diameter (TCD), and cisterna magnum (CM). cisterna magnum (CM). Fetal Intelligent Navigation Echocardiography (FINE) applied on a STIC volume us- Fetal Intelligent Navigation Echocardiography (FINE) applied on a STIC volume ing “intelligent navigation” technology allows the automatic display of nine standard fe- using “intelligent navigation” technology allows the automatic display of nine standard fetaltal echocardiography echocardiography views views [71]. [71 ].This This can can simplify simplify fetal fetal cardiac cardiac examinations, examinations, reduce reduce op- operatorerator dependency, dependency, and and help help detect detect congenital congenital heart heart defects defects [71]. [71]. DuringDuring a a real-time real-time 2D 2D morphology morphology scan, scan, identifying identifying and and interpreting interpreting fetal fetal standard standard scanscan planes planes are are highly highly complex complex tasks. tasks. With Wi automaticth automatic image processingimage processing technology technology [46,70], these[46,70], tasks these can tasks be assisted can be by assisted providing by providing feedback orfeedback guidance or toguidance an ultrasound to an ultrasound operator onoperator whether on a whether correct standard a correct scan standard plane ofscan fetal plane anatomy of fetal is obtained,anatomy is whether obtained, all partswhether of anatomyall parts areof anatomy checked, andare checked, whether unusualand whether findings unusual on a standard findings planeon a standard are identified plane [72 are]. Theidentified operator [72]. can The use thisoperator technology can use as this a second technology pass or as confirmation a second pass to improve or confirmation diagnostic to accuracyimprove [ diagnostic70]. This can accuracy also allow [70]. audit This can and al qualityso allow improvement audit and quality [73]. improvement [73]. BasedBased on on deep deep learning, learning, image image segmentation segmentation is is an an image image processing processing method method that that can can automaticallyautomatically recognizerecognize thethe location location and and size size of of an an object object in in pixels. pixels. However, However, accurate accurate segmentationsegmentation of of most most anatomical anatomical structures structures in in medical medical ultrasound ultrasound is is limited limited by by the the low low contrastcontrast between between the the target target and and background background of of the the images images [74 [74].]. To To improve improve the the segmenta- segmen- tiontation performance performance of of the the thoracic thoracic wall wall in in fetal fetal ultrasound ultrasound videos, videos, a a novel novel model-agnostic model-agnostic methodmethod using using deep deep learning learning techniques techniques in processing in processing time-series time-series information information of ultrasound of ultra- videossound andvideos the and shape the of shape the thoracic of the thoracic wall was wall proposed was proposed [75]. Accurate [75]. Accurate segmentation segmenta- can assisttion ultrasonographerscan assist ultrasonographers with identifying with theidenti thoracicfying areathe thoracic and its orientation, area and its and orientation, it has the potentialand it has to the build potential AI-based to build diagnostic AI-based support diagnostic models support to assess models four-chamber to assess view four-cham- [75]. ber Thereview [75]. are emerging studies on the application of artificial intelligence in obstetric scan. ItThere is feasible are emerging to use 3D studies ultrasound on the to application automatic measureof artificial thymic intelligence volume in [76 obstetric]. Two- dimensionalscan. It is feasible (2D) placental to use 3D sonographic ultrasound to images automatic can be measure screened thymic for lacunae, volume which [76]. Two- is a featuredimensional of PAS (2D) [77]. Preliminaryplacental sono resultsgraphic are encouraging.images can be Further screened improvement for lacunae, of algorithmwhich is a and technology are required prior to using AI applications in clinical practice.

8. Conclusions The use of high-resolution ultrasonography can facilitate detailed diagnostic ultrasonography, in particular, fetal echocardiography and targeted neurosonography, in at-risk pregnancies, as suggested by the recent guidelines. The use of radiant ﬂow can improve the display, especially in complex cardiac or vascular structures. The use of 3D/4D ul- Diagnostics 2021, 11, 1217 15 of 18

trasound may help in the prenatal diagnosis and counseling of some fetal abnormalities. Select use of linear transducer may enhance the diagnostic capabilities of some superficial anomalies. Speckle tracking of the fetal heart can allow assessment of fetal heart shape, size and contractibility, and further studies are required to assess its clinical effectiveness. At present, automated tools for simple task such as measurement of fetal growth biometry are a good assistant to routine ultrasonography. Further refinement of automated algorithm is required, especially for complex tasks, to improve the workflow.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/diagnostics11071217/s1. Video S1. High-resolution ultrasonography of the fetal heart at 20 weeks’ gestation showing four-chamber view with two pulmonary veins connecting to the left atrium at a high frame rate. Video S2. High-resolution ultrasonography of the fetal brain at 20 weeks’ gestation: (a) mid-sagittal view showing corpus callosum, thalamus, brain stem and cerebellar vermis, and (b) coronal view through anterior fontanelle showing interhemispheric fissure, corpus callosum and cavum septi pellcidi, thalami. Video S3. High-resolution ultrasonography of the fetal face at 20 weeks’ gestation: (a) mid-sagittal view showing facial profile, ‘superimposed line’, opening and closing of mouth, (b) transverse view showing orbits, lips, palate, and maxilla. Video S4. High- resolution ultrasonography of the fetal neck at 21 weeks gestation: coronal view showing abduction and adduction of vocal cords. Video S5. High-resolution ultrasound of the fetus at 12–13 weeks’ gestation: (a) mid-sagittal view showing head, neck and facial profile, (b) coronal view showing both eyes and ears, (c) four-chamber heart at a high frame rate, and (d) color flow imaging showing blood flow at four-chamber, three-vessel and right subclavian artery views. Video S6. Radiant flow imaging of the fetal heart at 20 weeks’ gestation: (a) Four-chamber view and bifurcation of the pulmonary artery, and (b) three-vessel trachea view and right subclavian artery. Radiant flow shows the blood flow with a sense of depth. Video S7. Radiant flow imaging of the fetal brain at 20 weeks’ gestation showing callosum artery. Radiant flow shows the blood flow with a sense of depth. Video S8. Radiant flow imaging of the fetal umbilical–portal venous system at 20 weeks’ gestation: (a) transverse view of the abdomen showing umbilical vein, left portal vein, and right portal vein, ductus venosus and main umbilical vein, and (b) sagittal view showing the merging of ductus venosus, left hepatic vein, inferior vena cava into the right atrium of the heart. Radiant flow shows the blood flow with a sense of depth. Video S9. A cine loop of glass-body mode of a spatio-temporal image correlation (STIC) volume in color Doppler showing a rendered image of a normal fetal heart in a cardiac cycle at 20 weeks’ gestation. Video S10. A cine loop of three-dimensional rendered images of the fetal long bones at 20–22 weeks’ gestation: (a) upper limb including humerus, ulnar, radius and five fingers, and (b) upper and lower limbs. (c) A video clip of four-dimensional rendered image of fetal spine at 20 weeks’ gestation. Funding: This review received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Permission of all figures was obtained from the patients. Data Availability Statement: Data is contained within the article or Supplementary Materials. Conflicts of Interest: The author declares no conflict of interest.

References 1. Salomon, L.J.; Alﬁrevic, Z.; Berghella, V.; Bilardo, C.; Hernandez-Andrade, E.; Johnsen, S.L.; Kalache, K.; Leung, K.Y.; Malinger, G.; Munoz, H.; et al. ISUOG Clinical Standards Committee. Practice guidelines for performance of the routine mid-trimester fetal ultrasound scan. Ultrasound Obstet. Gynecol. 2011, 37, 116–126. [CrossRef][PubMed] 2. Anonymous. AIUM Practice Parameter for the Performance of Detailed Second- and Third-Trimester Diagnostic Obstetric Ultrasound Examinations. J. Ultrasound Med. 2019, 38, 3093–3100. [CrossRef][PubMed] 3. Anonymous. AIUM Practice Parameter for the Performance of Fetal Echocardiography. J. Ultrasound Med. 2020, 39, E5–E16. 4. Malinger, G.; Paladini, D.; Haratz, K.K.; Monteagudo, A.; Pilu, G.L.; Timor-Tritsch, I.E. ISUOG Practice Guidelines (updated): Sonographic examination of the fetal central nervous system. Part 1: Performance of screening examination and indications for targeted neurosonography. Ultrasound Obstet. Gynecol. 2020, 56, 476–484. [CrossRef] 5. Paladini, D.; Malinger, G.; Birnbaum, R.; Monteagudo, A.; Pilu, G.; Salomon, L.J.; Timor-Tritsch, I.E. ISUOG Practice Guidelines (updated): Sonographic examination of the fetal central nervous system. Part 2: Performance of targeted neurosonography. Ultrasound Obstet. Gynecol. 2021, 57, 661–671. [CrossRef] Diagnostics 2021, 11, 1217 16 of 18

6. Salomon, L.J.; Alfirevic, Z.; Audibert, F.; Kagan, K.O.; Paladini, D.; Yeo, G.; Raine-Fenning, N.; ISUOG Clinical Standards Committee. ISUOG consensus statement on the impact of non-invasive prenatal testing (NIPT) on prenatal ultrasound practice. Ultrasound Obstet. Gynecol. 2014, 44, 122–123. [CrossRef] 7. Kenkhuis, M.J.A.; Bakker, M.; Bardi, F.; Fontanella, F.; Bakker, M.K.; Fleurke-Rozema, J.H.; Bilardo, C.M. Effectiveness of a 12-13-week scan for the early diagnosis of fetal congenital anomalies in the cell-free DNA era. Ultrasound Obstet. Gynecol. 2018, 51, 463–469. [CrossRef] 8. Ficara, A.; Syngelaki, A.; Hammami, A.; Akolekar, R.; Nicolaides, K.H. Value of routine ultrasound examination at 35–37 weeks’ gestation in diagnosis of fetal abnormalities. Ultrasound Obstet. Gynecol. 2020, 55, 75–80. [CrossRef] 9. Edwards, L.; Hui, L. First and second trimester screening for fetal structural anomalies. Semin. Fetal. Neonatal. Med. 2018, 23, 102–111. [CrossRef] 10. International Society of Ultrasound in Obstetrics and Gynecology; Carvalho, J.S.; Allan, L.D.; Chaoui, R.; Copel, J.A.; DeVore, G.R.; Hecher, K.; Lee, W.; Munoz, H.; Paladini, D.; et al. ISUOG Practice Guidelines (updated): Sonographic screening examination of the fetal heart. Ultrasound Obstet. Gynecol. 2013, 41, 348–359. [CrossRef] 11. Platt, L.D. Should the first trimester ultrasound include anatomy survey? Semin. Perinatol. 2013, 37, 310–322. [CrossRef][PubMed] 12. Pooh, R.K.; Kurjak, A. 3D/4D sonography moved prenatal diagnosis of fetal anomalies from the second to the first trimester of pregnancy. J. Matern. Fetal. Neonatal. Med. 2012, 25, 433–455. [CrossRef][PubMed] 13. Paladini, D.; Vassallo, M.; Tartaglione, A.; Lapadula, C.; Martinelli, P. The role of tissue harmonic imaging in fetal echocardiography. Ultrasound Obstet. Gynecol. 2004, 23, 159–164. [CrossRef][PubMed] 14. Biard, J.-M.; Steenhaut, P.; Bernard, P.; Race, V.; Sznajer, Y. Antenatal diagnosis of cardio-facio-cutaneous syndrome: Prenatal characteristics and contribution of fetal facial dysmorphic signs in utero. About a case and review of literature. Eur. J. Obstet. Gynecol. Reprod. Biol. 2019, 240, 232–241. [CrossRef][PubMed] 15. Mak, A.S.L.; Ma, T.W.L.; Chan, K.Y.K.; Kan, A.S.Y.; Tang, M.H.Y.; Leung, K.Y. Prenatal diagnosis of 5p deletion syndrome: Report of five cases. J. Obstet. Gynaecol. Res. 2019, 45, 923–926. [CrossRef] 16. Faure, J.M.; Mousty, E.; Bigorre, M.; Wells, C.; Boulot, P.; Captier, G.; Fuchs, F. Prenatal ultrasound diagnosis of cleft palate without cleft lip, the new ultrasound semiology. Prenat. Diagn. 2020, 40, 1447–1458. [CrossRef] 17. Yu, F.; Leung, K.Y. Prenatal sonographic appearance of laryngeal atresia: A case report. J. Clin. Ultrasound. 2020, 48, 244–246. [CrossRef] 18. Rogelio, C.M.; Alma, G.V.; Miguel, M.R.; Jonahtan, L.G.; Hugo, L.B.; Eréndira, C.G.; Israel, J.M.; Areli, R.N.; Rosa, V.G.; César, F.H. Prenatal diagnosis of laryngo-tracheo-esophageal anomalies in fetuses with congenital diaphragmatic hernia by ultrasound evaluation of the vocal cords and fetal laryngoesophagoscopy. Prenat. Diagn. 2020, 40, 1540–1546. 19. Yu, F.N.Y.; Leung, K.Y. Diagnosis and prediction of miscarriage: Can we do better? Hong Kong Med. J. 2020, 26, 90–92. [CrossRef] 20. Dooley, W.M.; Braud, L.D.; Thanatsis, N.; Memtsa, M.; Jauniaux, E.; Jurkovic, D. Predictive value of presence of amniotic sac without visible embryonic heartbeat in diagnosis of early embryonic demise. Ultrasound Obstet. Gynecol. 2021, 57, 149–154. [CrossRef] 21. Nadim, B.; Infante, F.; Lu, C.; Sathasivam, C.; Condous, G. Morphological ultrasound types known as ‘blob’ and ‘bagel’ signs should be reclassified from suggesting probable to indicating definite tubal ectopic pregnancy. Ultrasound Obstet. Gynecol. 2018, 51, 543–549. [CrossRef][PubMed] 22. Yu, F.N.Y.; Leung, K.Y. Antenatal diagnosis of placenta accreta spectrum (PAS) disorders. Best Pract. Res. Clin. Obstet. Gynaecol. 2021, 72, 13–24. [CrossRef][PubMed] 23. D’Antonio, F.; Timor-Tritsch, I.E.; Palacios-Jaraquemada, J.; Monteagudo, A.; Buca, D.; Forlani, F.; Minneci, G.; Foti, F.; Manzoli, L.; Liberati, M.; et al. First-trimester detection of abnormally invasive placenta in high-risk women: Systematic review and meta-analysis. Ultrasound Obstet. Gynecol. 2018, 51, 176–183. [CrossRef] 24. Salomon, L.J.; Alfirevic, Z.; Bilardo, C.M.; Chalouhi, G.E.; Ghi, T.; Kagan, K.O.; Lau, T.K.; Papageorghiou, A.T.; Raine-Fenning, N.J.; Stirnemann, J.; et al. ISUOG practice guidelines: Performance of first-trimester fetal ultrasound scan. Ultrasound Obstet. Gynecol. 2013, 41, 102–113. [PubMed] 25. Anonymous. AIUM Practice Parameter for the Performance of Detailed Diagnostic Obstetric Ultrasound Examinations between 12 Weeks 0 Days and 13 Weeks 6 Days. J. Ultrasound Med. 2021, 40, E1–E16. 26. Minnella, G.P.; Crupano, F.M.; Syngelaki, A.; Zidere, V.; Akolekar, R.; Nicolaides, K.H. Diagnosis of major heart defects by routine first trimester ultrasound examination: Association with increased nuchal translucency, tricuspid regurgitation and abnormal flow in ductus venosus. Ultrasound Obstet. Gynecol. 2020, 55, 637–644. [CrossRef] 27. Bhide, A.; Acharya, G.; Bilardo, C.M.; Brezinka, C.; Cafici, D.; Hernandez-Andrade, E.; Kalache, K.; Kingdom, J.; Kiserud, T.; Lee, W.; et al. ISUOG practice guidelines: Use of Doppler ultrasonography in obstetrics. Ultrasound Obstet. Gynecol. 2013, 41, 233–239. [PubMed] 28. Martinez-Portilla, R.J.; Caradeux, J.; Meler, E.; Lip-Sosa, D.L.; Sotiriadis, A.; Figueras, F. Third-trimester uterine artery Doppler for prediction of adverse outcome in late small-for-gestational-age fetuses: Systematic review and meta-analysis. Ultrasound Obstet. Gynecol. 2020, 55, 575–585. [CrossRef] 29. Sirico, A.; Diemert, A.; Glosemeyer, P.; Hecher, K. Prediction of adverse perinatal outcome by cerebroplacental ratio adjusted for estimated fetal weight. Ultrasound Obstet. Gynecol. 2018, 51, 381–386. [CrossRef] Diagnostics 2021, 11, 1217 17 of 18

30. Khalil, A.; Morales-Roselló, J.; Townsend, R.; Morlando, M.; Papageorghiou, A.; Bhide, A.; Thilaganathan, B. Value of third-trimester cerebroplacental ratio and uterine artery Doppler indices as predictors of stillbirth and perinatal loss. Ultrasound Obstet. Gynecol. 2016, 47, 74–80. [CrossRef] 31. MacDonald, T.M.; Hui, L.; Robinson, A.J.; Dane, K.M.; Middleton, A.L.; Tong, S.; Walker, S.P. Cerebral-placental-uterine ratio as novel predictor of late fetal growth restriction: Prospective cohort study. Ultrasound Obstet. Gynecol. 2019, 54, 367–375. [CrossRef] 32. Ghi, T.; Eggebø, T.; Lees, C.; Kalache, K.; Rozenberg, P.; Youssef, A.; Salomon, L.J.; Tutschek, B. ISUOG Practice Guidelines: Intrapartum ultrasound. Ultrasound Obstet. Gynecol. 2018, 52, 128–139. [CrossRef] 33. Hjartardóttir, H.; Lund, S.H.; Benediktsdóttir, S.; Geirsson, R.T.; Eggebø, T.M. Can ultrasound on admission in active labor predict labor duration and a spontaneous delivery? Am. J. Obstet. Gynecol. MFM. 2021.[CrossRef][PubMed] 34. Hjartardóttir, H.; Lund, S.H.; Benediktsdóttir, S.; Geirsson, R.T.; Eggebø, T.M. Fetal descent in nulliparous women assessed by ultrasound: A longitudinal study. Am. J. Obstet. Gynecol. 2021, 224, 378.e1–378.e15. [CrossRef][PubMed] 35. Kahrs, B.H.; Usman, S.; Ghi, T.; Youssef, A.; Torkildsen, E.A.; Lindtjørn, E.; Østborg, T.B.; Benediktsdottir, S.; Brooks, L.; Harmsen, L.; et al. Sonographic prediction of outcome of vacuum deliveries: A multicenter, prospective cohort study. Am. J. Obstet. Gynecol. 2017, 217, 69.e1–69.e10. [CrossRef] 36. Usman, S.; Wilkinson, M.; Barton, H.; Lees, C.C. The feasibility and accuracy of ultrasound assessment in the labor room. J. Matern. Fetal Neonatal. Med. 2019, 32, 3442–3451. [CrossRef][PubMed] 37. Wang, B.; Zhang, Y.; Chen, S.S.; Xiang, X.W.; Wen, J.; Yi, M.; He, B.Y.; Hu, B. Diagnostic accuracy of cervical elastography in predicting preterm delivery: A systematic review and meta-analysis. Meta-Anal. Med. 2019, 98, e16449. [CrossRef] 38. Lu, J.; Cheng, Y.K.Y.; Ho, S.Y.S.; Sahota, D.S.; Hui, L.L.; Poon, L.C.; Leung, T.Y. The predictive value of cervical shear wave elastography in the outcome of labor induction. Acta Obstet. Gynecol. Scand. 2020, 99, 59–68. [CrossRef] 39. Mak, A.S.L.; Leung, K.Y. Prenatal ultrasonography of craniofacial abnormalities. Ultrasonography 2019, 38, 13–24. [CrossRef] 40. Persico, N.; Moratalla, J.; Lombardi, C.M.; Zidere, V.; Allan, L.; Nicolaides, K.H. Fetal echocardiography at 11-13 weeks by transabdominal high-frequency ultrasound. Ultrasound Obstet. Gynecol. 2011, 37, 296–301. [CrossRef] 41. Votino, C.; Kacem, Y.; Dobrescu, O.; Dessy, H.; Cos, T.; Foulon, W.; Jani, J. Use of a high-frequency linear transducer and MTI ﬁltered color ﬂow mapping in the assessment of fetal heart anatomy at the routine 11 to 13 + 6-week scan: A randomized trial. Ultrasound Obstet. Gynecol. 2012, 39, 145–151. [CrossRef] 42. Wang, Y.; Zhang, Y. Fetal Vascular Rings and Pulmonary Slings: Strategies for Two- and Three-Dimensional Echocardiographic Diagnosis. J. Am. Soc. Echocardiogr. 2021, 34, 336–351. [CrossRef][PubMed] 43. Yin, H.; Li, Y. Diagnostic value of Omniview technique on the agenesis of corpus callosum. J. Matern. Fetal. Neonatal. Med. 2019, 32, 3806–3811. [CrossRef][PubMed] 44. Wang, Y.; Zhang, Y.; Wang, M. Prenatal diagnosis of an aberrant ductus venosus draining into the coronary sinus using two- and three-dimensional echocardiography: A case report. BMC Pregnancy Childbirth. 2021, 21, 392. [CrossRef] 45. Merz, E.; Pashaj, S. Advantages of 3D ultrasound in the assessment of fetal abnormalities. J. Perinat. Med. 2017, 45, 643–650. [CrossRef] 46. Chaoui, R.; Abuhamad, A.; Martins, J.; Heling, K.S. Recent development in three- and four-dimension fetal echocardiography. Fetal. Diagn Ther. 2020, 47, 345–353. [CrossRef][PubMed] 47. Paladini, D.; Volpe, P. Posterior fossa and vermian morphometry in the characterization of fetal cerebellar abnormalities: A prospective three-dimensional ultrasound study. Ultrasound Obstet. Gynecol. 2006, 27, 482–489. [CrossRef][PubMed] 48. Pashaj, S.; Merz, E. Detection of Fetal Corpus Callosum Abnormalities by Means of 3D Ultrasound. Ultraschall Med. 2016, 37, 185–194. [CrossRef][PubMed] 49. Frisova, V.; Srutova, M.; Hyett, J. 3-D Volume Assessment of the Corpus Callosum and Cerebellar Vermis Using Various Volume Acquisition and Post-Processing Protocols. Fetal. Diagn Ther. 2018, 43, 199–207. [CrossRef] 50. Wang, L.M.; Leung, K.Y.; Tang, M. Prenatal evaluation of facial clefts by three-dimensional extended imaging. Prenat. Diagn. 2007, 27, 722–729. [CrossRef] 51. Rotten, D.; Levaillant, J.M.; Martinez, H.; Pointe, H.D.L.; Vicaut, E. The fetal mandible: A 2D and 3D sonographic approach to the diagnosis of retrognathia and micrognathia. Ultrasound Obstet. Gynecol. 2002, 19, 122–130. [CrossRef][PubMed] 52. Tutschek, B.; Blaas, H.K.; Abramowicz, J.; Baba, K.; Deng, J.; Lee, W.; Merz, E.; Platt, L.; Pretorius, D.; Timor-Tritsch, I.E.; et al. ISUOG 3D Special Interest Group. Three-dimensional ultrasound imaging of the fetal skull and face. Ultrasound Obstet. Gynecol. 2017, 50, 7–16. [CrossRef] 53. Schut, P.C.; Cohen-Overbeek, T.E.; Galis, F.; Broek, C.M.A.T.; Steegers, E.A.; Eggink, A.J. Adverse Fetal and Neonatal Outcome and an Abnormal Vertebral Pattern: A Systematic Review. Obstet. Gynecol. Surv. 2016, 71, 741–750. [CrossRef][PubMed] 54. Helfer, T.M.; Peixoto, A.B.; Tonni, G.; Júnior, E.A. Craniosynostosis: Prenatal diagnosis by 2D/3D ultrasound, magnetic resonance imaging and computed tomography. Med. Ultrason. 2016, 18, 378–385. [CrossRef] 55. Dall’Asta, A.; Grisolia, G.; Nanni, M.; Volpe, N.; Schera, G.B.L.; Frusca, T.; Ghi, T. Sonographic demonstration of fetal esophagus using three-dimensional ultrasound imaging. Ultrasound Obstet. Gynecol. 2019, 54, 746–751. [CrossRef][PubMed] 56. Calì, G.; Foti, F.; Minneci, G. 3D power Doppler in the evaluation of abnormally invasive placenta. J. Perinat Med. 2017, 45, 701–709. [CrossRef][PubMed] 57. Tutschek, B.; Blaas, H.K. A human embryo in the palm of your hand. Ultrasound Obstet. Gynecol. 2017, 50, 539–540. [CrossRef] Diagnostics 2021, 11, 1217 18 of 18

58. Coté, J.J.; Badura-Brack, A.S.; Walters, R.W.; Dubay, N.G.; Bredehoeft, M.R. Randomized Controlled Trial of the Effects of 3D-Printed Models and 3D Ultrasonography on Maternal-Fetal Attachment. J. Obstet. Gynecol. Neonatal. Nurs. 2020, 49, 190–199. [CrossRef][PubMed] 59. Miller, J.L.; Ahn, E.S.; Garcia, J.R.; Miller, G.T.; Satin, A.J.; Baschat, A.A. Ultrasound-based three-dimensional printed medical model for multispecialty team surgical rehearsal prior to fetoscopic myelomeningocele repair. Ultrasound Obstet. Gynecol. 2018, 51, 836–840. [CrossRef] 60. Chen, S.A.; Ong, C.S.; Hibino, N.; Baschat, A.A.; Garcia, J.R.; Miller, J.L. 3D printing of fetal heart using 3D ultrasound imaging data. Ultrasound Obstet. Gynecol. 2018, 52, 808–809. [CrossRef] 61. Veronese, P.; Bertelli, F.; Cattapan, C.; Andolfatto, M.; Gervasi, M.T.; Vida, V.L. Three-Dimensional Printing of Fetal Heart With d-Transposition of the Great Arteries From Ultrasound Imaging Data. World J. Pediatr. Congenit. Heart Surg. 2021, 12, 291–292. [CrossRef][PubMed] 62. DeVore, G.R.; Klas, B.; Satou, G.; Sklansky, M. 24-segment sphericity index: A new technique to evaluate fetal cardiac diastolic shape. Ultrasound Obstet. Gynecol. 2018, 51, 650–658. [CrossRef][PubMed] 63. DeVore, G.R.; Satou, G.; Sklansky, M. Abnormal Fetal Findings Associated With a Global Sphericity Index of the 4-Chamber View Below the 5th Centile. J. Ultrasound Med. 2017, 36, 2309–2318. [CrossRef][PubMed] 64. Hobbins, J.C.; Gumina, D.L.; Zaretsky, M.V.; Driver, C.; Wilcox, A.; DeVore, G.R. Size and shape of the four-chamber view of the fetal heart in fetuses with an estimated fetal weight less than the tenth centile. Am. J. Obstet. Gynecol. 2019, 221, e1–e495. [CrossRef] 65. Van Oostrum, N.H.; Chantelle, M.; van der Woude, D.A.; Kemps, H.M.; Oei, S.G.; van Laar, J.O. Fetal strain and strain rate during pregnancy measured with speckle tracking echocardiography: A systematic review. Eur. J. Obstet. Gynecol. Reprod. Biol. 2020, 250, 178–187. [CrossRef] 66. Van Oostrum, N.H.; Derks, K.; van der Woude, D.A.; Clur, S.A.; Oei, S.G.; van Laar, J.O. Two-dimensional Speckle tracking echocardiography in Fetal Growth Restriction: A systematic review. Eur. J. Obstet. Gynecol. Reprod. Biol. 2020, 254, 87–94. [CrossRef] 67. Kim, H.P.; Lee, S.M.; Kwon, J.Y.; Park, Y.; Kim, K.C.; Seo, J.K. Automatic evaluation of fetal head biometry from ultrasound images using machine learning. Physiol. Meas. 2019, 40, 065009. [CrossRef] 68. Salim, I.; Cavallaro, A.; Ciofolo-Veit, C.; Rouet, L.; Raynaud, C.; Mory, B.; Billon, A.C.; Harrison, G.; Roundhill, D.; Papageorghiou, A.T. Evaluation of automated tool for two-dimensional fetal biometry. Ultrasound Obstet. Gynecol. 2019, 54, 650–654. [CrossRef] 69. Society of Diagnostic Medical Sonography. Industry Standards for the Prevention of Work Related Musculoskeletal Disorders in Sonography. J. Diagn. Med. Sonogr. 2017, 33, 371–391. 70. Pluym, I.D.; Afshar, Y.; Holliman, K.; Kwan, L.; Bolagani, A.; Mok, T.; Silver, B.; Ramirez, E.; Han, C.S.; Platt, L.D. Accuracy of automated three-dimensional ultrasound imaging technique for fetal head biometry. Ultrasound Obstet. Gynecol. 2021, 57, 798–803. [CrossRef] 71. Yeo, L.; Romero, R. New and advanced features of fetal intelligent navigation echocardiography (FINE) or 5D heart. J. Matern. Fetal Neonatal. Med. 2020, 6, 1–19. [CrossRef] 72. Drukker, L.; Noble, J.A.; Papageorghiou, A.T. Introduction to artiﬁcial intelligence in ultrasound imaging in obstetrics and gynecology. Ultrasound Obstet. Gynecol. 2020, 56, 498–505. [CrossRef][PubMed] 73. Abramowicz, J.S. Obstetric ultrasound: Where are we and where are we going? Ultrasonography 2021, 40, 57–74. [CrossRef] [PubMed] 74. Liu, S.; Wang, Y.; Yang, X.; Lei, B.; Liu, L.; Li, S.X.; Ni, D.; Wang, T. Deep Learning in Medical Ultrasound Analysis: A Review. Engineering 2019, 5, 261–275. [CrossRef] 75. Shozu, K.; Komatsu, M.; Sakai, A.; Komatsu, R.; Dozen, A.; Machino, H.; Yasutomi, S.; Arakaki, T.; Asada, K.; Kaneko, S.; et al. Model-Agnostic Method for Thoracic Wall Segmentation in Fetal Ultrasound Videos. Biomolecules 2020, 10, 1691. [CrossRef] [PubMed] 76. Re, C.; Bertucci, E.; Weissmann-Brenner, A.; Achiron, R.; Mazza, V.; Gindes, L. Fetal thymus volume estimation by virtual organ computer-aided analysis in normal pregnancies. J. Ultrasound Med. 2015, 34, 847–852. [CrossRef] 77. Qi, H.; Collins, S.; Noble, J.A. Automatic Lacunae Localization in Placental Ultrasound Images via Layer Aggregation. Med. Image Comput. Comput. Assist. Interv. 2018, 11, 921–929.