Manuscript Details

Manuscript number EHD_2018_353

Title 3D EVALUATION OF FETAL BRAIN STRUCTURES: REFERENCE VALUESAND CLINICAL SIGNIFICANCE

Article type Research Paper

Abstract

Abstract Background The development of the fetal central nervous system is one of the most important fields ofresearch in perinatology. Since the early 1980s, 3D ultrasound has become one of the major research tools inobstetrics and gynecology, Aims The first aim of this study was to reconstruct thalamus, cerebellum and hemi-cortexvolumes of fetal brain and generate, for these volumes, growth curve evaluating the correlation between volumetricparameters with gestational age and fetal growth. The second aim was to evaluate the presence of any significantdifference between brain volumes of fetuses with IUGR and fetuses of diabetic mothers versus fetuses ofuncomplicated pregnancies (controls). Study design Over 3 years, we enrolled 368 pregnant women. Using“Tomographic Ultrasound Imaging” (TUI), in all cases we obtained a satisfying 3D acquisition of fetal brain. Wereconstructed offline thalamus, cerebellum and hemi-cortex volumes using “Virtual Organ Computer-aidedAnaLysis” (VOCAL) or 4D View. Results Among the 361 fetuses examined, we obtained 339 thalamus volumes, 275cerebellum volumes and 275 hemi-cortex volumes. We compared the volumes of the fetal brain structures inuncomplicated pregnancies (controls) with the volumes of fetuses in diabetic mothers (15 patients) and of IUGRfetuses (34 patients) finding statistically significant differences. Conclusion Our study confirms the reliability of cerebralvolumes evaluation using 3D technology and how they grow through gestation. Furthermore, we found a differentgrowth capacity of fetal brain structures exposed to specific injuries like hyperglycemia and hypoxia.

Keywords Keywords: Thalamus, Cerebellum, TUI, VOCAL, Neurosonography, Fetal Brain

Corresponding Author Giulia Babucci

Corresponding Author'sInstitution

CEMER (European Medical and Research Centre) via del Giochetto 53/A06134, Perugia, Italy Tel. +39 075 9287355 Cell. +39 340 8244019

Order of Authors Giulia Babucci, Gian DI Renzo, Benito Cappuccini, Karl Rosen, Graziano Clerici

Suggested reviewers Gerard Visser, Karel Marsal, Domenico Arduini

Submission Files Included in this PDF

File Name [File Type]

Cover letter.docx [Cover Letter]

Highlights.docx [Highlights]

Title page.docx [Title Page (with Author Details)]

Main Tex.docx [Manuscript (without Author Details)]

Figure.docx [Figure]

Author Statements.docx [Author Statement]

capture.docx [Supplementary Material]

Refereces.docx [Supplementary Material]

To view all the submission files, including those not included in the PDF, click on the manuscript title on your EVISEHomepage, then click 'Download zip file'.

Research Data Related to this Submission

There are no linked research data sets for this submission. The following reason is given:Data are bigger, for each patient it was collected more image of this structures and after it was calcuated andreconstructed volume. It's no possible to upload all work data

To Professor E. F. MaaloufEditor-in-Chief, Early Human Development

November 5, 2018

Dear Professor Maalouf,I am pleased to submit an original research article entitled “3D EVALUATION OF FETAL BRAIN STRUCTURES: REFERENCE VALUES AND CLINICAL SIGNIFICANCE” by Giulia Babucci, Gian Carlo Di Renzo, Benito Cappuccini, Karl Gustaf Rosén, Graziano Clerici, for publication in the Journal of Early Human Development.This study shows the role of 3D ultrasound technology in the evaluation of fetal brain structures and the possible consequences of the exposition of these structures to specific injuries like hyperglycemia and oxygen deficiency.This manuscript has not been published and is not under consideration for publication elsewhere. We have no conflicts of interest to disclose.

Thank you for your consideration!

Sincerely,

Giulia Babucci, MDEuropean Medical and Research Center (CEMER)via del Giochetto 53/A06134, Perugia, Italygiuliababucci@gmail.comtel. +39 340 8244019

Our study confirms the reliability of 3D ultrasonography in evaluating the volume of fetal brain and the growth of its structures through gestation. This is demonstrated in our growth curves using percentiles of fetal growth. Previous studies realized growth curves of fetal brain structures using MRI. Our study is the first using ultrasound technique with, to our knowledge, the largest number of cases ever enrolled.

Unlike the previous data, showing lack of homogeneity in the growth response of different brain structures subjected to hypoxic insult, in our study the IUGR fetuses showed a tendency to increase the volume of all the brain structures examined. This enhanced growth became statistically significant for both the thalamus and the hemi-cortex. We can speculate that this finding is related to the vasodilatation of the cerebral circulation due to oxygen deficiency in IUGR fetuses. The relative greater blood supply, in the cerebral area, due to the reactive vasodilatation, would be required to maintain oxygen supply and enhanced growth. Possibly, this could also be related to the edema of these cerebral structures, due to the increase of vascular permeability related to the vasodilation and to the release of pro-inflammatory vasoactive substances in response to hypoxia.

In conclusion, our study suggests that it’s possible to obtain fetal cerebral structures volumetric growth curves by 3D ultrasound examination, investigating fetuses between 17 and 39 weeks gestation. The use of tissue volumes for the construction of growth curves could add important informations to the biometric 2D curve. Furthermore, this study suggests that the fetal brain is influenced in different ways, when exposed to a specific insult (like oxygen deficiency in IUGR and hyperglycemia in diabetes). For those reasons the fetal brain can be considered a dynamic structure that changes its response depending on the type, amount and characteristics of the injury

3D EVALUATION OF FETAL BRAIN STRUCTURES: REFERENCE VALUES AND CLINICAL SIGNIFICANCE

Giulia Babucci MD; Gian Carlo Di Renzo, MD PhD; Benito Cappuccini, MD; Karl Gustaf Rosén, MD; Graziano Clerici, MD PhD

Corresponding Author:

Babucci Giulia, MDCEMER (European Medical and Research Centre)via del Giochetto 53/A06134, Perugia, ItalyTel. +39 075 9287355Cell. +39 340 8244019giuliababucci@gmail.com

Coauthors:

Di Renzo GianCarlo, MD PhD Ospedale Santa Maria della Misericordia06100, Perugia, Italygiancarlo.direnzo@unipg.it

Benito Cappuccini, MDOspedale Santa Maria della Misericordia06100, Perugia, Italybenitocappuccini@hotmail.com

Karl Gustaf Rosén, MD Department of Physiology, Sahlgren’s AcademyUniversity of GothenburgGothenburg, Swedenkg.rosen@neoventor.se

Graziano Clerici, MD PhDSechenov UniversityTrubetskaya str. 8b2119992, Moscow, Russiagraziano.clerici@gmail.com

Introduction

The development of the fetal central nervous system (CNS) is one of the most important fields of research in perinatology. Since the early 1980s, 3D ultrasound, and later neurosonography, have become research tools in obstetrics and gynecology but they have not been incorporated into routine fetal anomaly scanning in most centers1,2. Several studies have shown that 3D ultrasonographic volume evaluation is accurate with an excellent intra-observer and inter-observer reliability3. Furthermore, previous studies have shown that these volumes can also be used for estimating gestational age4,5. Finally, the advantage of 3D ultrasonography over standard 2D scanning for the diagnosis of fetal anomalies has been reported3,6,7.

Of the CNS structures, thalamus, cerebellum and whole cerebrum are of special interest. Cerebellum is involved in a wide variety of functions such as the control of body movements and in neurocognitive functions8. Indeed, cerebellar lesions are associated with “cerebellar cognitive affective syndrome” described by Shmahmann9. The clinical interest for cerebellum developmental sequence is related to neuropsychiatric disorders, like autism and schizophrenia, which include cerebellar pathologies as part of their phenotype8. Thalami nuclei are important for a wide range of sensorimotor and neuropsychic functions as shown in thalamic vascular accidents10. Moreover, the disorders in thalamic development and volume are implicated in complex psychiatric syndromes and autism11. Cortex volume and brain volume were investigated in previous studies that showed how these structures are extremely sensitive to metabolic changes during intrauterine life12. Previous ultrasound studies of the brain cortex evaluated few cases and never tried to create a model of volumetric growth curves1-11. The first aim of this study was to realize the growth volumes curves for thalamus, cerebellum and fetal cerebrum, showing the correlation between volumetric parameters and gestational age. The second aim was to compare fetal brain volumes in IUGR fetuses and in fetuses of diabetic women versus controls.

Methods

A total of 361 pregnant women examined for routine obstetric ultrasound were included in this prospective observational study. The women’s gestational age ranged from 21 to 39 weeks and all cases were singleton pregnancies. Informed consent was obtained from each patient. All cases included in the study had a normal fetal brain anatomy and this aspect was confirmed after birth. Only cases without significant abnormal findings in a routine fetal anomaly scan were included in this study. As the following findings were not considered clinically relevant, we have enrolled two cases with isolated bovine aortic arch, five cases with mild pyelectasis, two cases with isolated single umbilical artery, 4 cases with isolated mild regurgitation of the tricuspid valve and ten cases with isolated hyperechogenic focus of the fetal heart. All pregnancies enrolled were correctly dated during the first trimester, by ultrasound. We excluded from the study all the patients for whom it was impossible to obtain a satisfactory 3D volume capture of the fetal head for unsuitable fetal position and/or unfavorable maternal habitus. Of 361 fetuses enrolled 34 were late onset IUGR

representing 9.4 % of the sample and 15 were fetuses of diabetic mothers representing the 4.1 % of the sample.

Ultrasound and sampling techniques:

All patients were scanned by two of the authors as trainer operators (G.B., G.C.) using a GE Voluson E10 ultrasound machine with 4-8 MHz curvilinear probe. Brain volumes were stored on digital device for further analysis. All volumes were acquired in axial view using the “Tomographic Ultrasound Imaging” (TUI) tool. The volume sample box was adjusted to include the complete fetal head and no magnification was used by zooming but adjusting depth. The angle volume was set at 80° and the highest quality acquisition was chosen. Brain volumes were acquired from each patient, at level of BPD plane. In many cases, this acquisition of volume allowed the sampling of the thalamus, cerebellum and unilateral cerebrum. In some cases, to avoid ultrasound shadowing of the petrous process and to obtain a clear image of cerebellum, a second volume data acquisition was necessary. This second acquisition was obtained from the same axial plane tilting the transducer of 15°-20° downward. Fetal volumes were acquired in absence of maternal and fetal movements. The multi-planar acquisition process was repeated until the established criteria were satisfied. From the trans-thalamic view and tilting the transducer downward, it was possible to evaluate the following brain structures: midline, both choroid plexus, lateral ventricles, III ventricle, thalami, inter-hemispheric fissure, cerebral falx, distal unilateral cerebrum, insula, posterior fossa with cerebellar hemispheres, vermis, IV ventricle, cisterna magna and cerebral peduncles.

Volume calculation was performed by a single operator using the “Virtual Organ Computer-aided Analysis” (VOCAL of GE Healthcare) or using the same application included in the 4D View software (GE Healthcare) for personal computer. As used in previous studies, we chose a rotation step of 30° obtaining 6 images of reconstruction volume. The rotation process, necessary to obtain volume, was started in an axial view, 0° of rotation, and finished in the same plane at 180° of rotation. In the middle (90° of rotation) we were able to obtain a coronal view of the structures. Because of the shadowing of the fetal skull at the ultrasound examination of the proximal structures, we evaluate the distal brain structures. In relation with the fetal position, we were able to evaluate the right brain structures in about half of cases and the left side in the other cases, finding not statistically significant differences in the detected volumes. Outlining the contours of the thalamus and cerebellum on six reconstruction planes we obtained a “Region Of Interest” (ROI) and thus a rendering with a volume expressed in cm3 (Fig. 1A, 1B, 2A, 2B). The same process was realized for unilateral cerebrum. For the same reasons described above, we evaluate the distal unilateral cerebrum that was easily recognizable (Fig. 3A, 3B). In about 25% of the cases examined the fetuses were actively moving and it was possible to evaluate both site of the fetal brain. Because we didn’t find statistically significant differences between the two hemi-structures, we decided to consider the global cerebrum volume doubling the volume of unilateral cerebrum.

Statistical Analysis

Statistical analysis was performed using SPSS software. We created specific growth curves for thalamus, cerebellum, unilateral cerebrum and total cerebral volumes. Given the inhomogeneity of the sample related to the gestational ages, we plotted the growth percentiles considered (5°, 25°,

50°, 75°, 95°) with nine classes of gestational age (class 1, 21+0; class 2, 21+1– 21+2; class 3, 21+3- 21+6; class 4, 21+6-22+3; class 5, 22+4-25+5; class 6, 25+6-27+0; class 7, 27+1-29+6; class 8, 30+0-33+1; class 9, 33+2-38+5). We used smoothing techniques and a second polynomial equation for the data processing. Then, we performed non-parametric comparisons, with Mann-Whitney test, for several variables: brain structures volumes, ratio between volumes of the different examined structures (unilateral cerebrum/thalamic - R1, unilateral cerebrum/cerebellar - R2 and thalamic/cerebellar - R3) and fetal weight at birth in the diabetic versus control group and in the IUGR versus control group. To describe the sample, we compared APGAR at 1 minute, at 5 minutes, growth centile and gestational age at birth between diabetic versus control group and between IUGR versus control group. P<0.05 was considered significant.

Results

Of the 361 fetuses enrolled in the study, we obtained 339 thalami volumes, 275 cerebellum volumes and 275 unilateral cerebrum volumes because it was impossible to obtain the volumes of all structures for each patient. The growth curve for the thalami volumes is shown in Figure 4, for the cerebellum volumes in Figure 5 and the growth curve for the unilateral cerebrum/total cerebrum volumes in Figures 6 and 7. For the cerebellar volume (Figure 5), we choose to consider only 8 classes for the small number of patients in the 9th class of gestational age. After the construction of the reference growth curves for uncomplicated pregnancies of the examined brain structures, we use these data as the reference for the comparison with fetuses with IUGR and fetuses of diabetic mothers. From statistical analysis, we found significantly larger thalamic volumes (p = 0.004) in the diabetic group than controls. For the ratio between structures (R1, R2, R3) and cerebellum volumes there were not significant differences between the diabetic group and the controls. About the variables “fetal weight at birth”, “APGAR score” and “centile of growth” we found not significant differences. Finally, the variable “gestational age at birth” was lower, as expected, in the diabetic group (p = 0.01) because of the incidence of the induction of labor. The same comparison with the same variables in the IUGR group versus the controls resulted in: significantly larger thalamic volumes in IUGR group versus control (p=0.019) and no statistically significant differences were observed for cerebellum but increased absolute values were detected in IUGR group than controls. Unilateral cerebrum volumes were significantly larger in the IUGR group than control (p = 0.032), while ratio between structures (R1, R2, R3) did not show significant variations between the IUGR group and the controls. As expected, fetal weight at birth and centile of growth were significantly lower in IUGR group. Finally, gestational age at the birth resulted significantly lower in the IUGR group than controls (p= 0.001).

Discussion

Our study confirms the reliability of 3D ultrasonography in the evaluation of cerebral volumes and the growth of these structures through gestation10,13,14. Previous studies realized growth curves of fetal brain structures using MRI15. Our study is the first using ultrasound technique with, to our knowledge, the largest number of cases ever enrolled. The absence of the 9th class for the cerebellum curve can be explained by the difficulties of sampling this cerebral structure due to the

shadow of the petrous process and the unfavorable fetal positions at later gestational ages. Unlike the previous data16,17,18, showing lack of homogeneity in the growth response of different brain structures subjected to hypoxic insult, in our study the IUGR fetuses showed a tendency to increase the volume of all the brain structures. This enhanced growth became statistically significant for both the thalamus and the unilateral cerebrum. The absolute value of the cerebellum volume was increased in IUGR group but did not achieve a statistical significance. In this case, the small number of the samples was probably responsible for the lack of statistical significance (34 cases of IUGR with 31 thalamic volumes and only 19 cerebellar volumes). We can speculate that this finding is related to the vasodilatation of the cerebral vascular district due to oxygen deficiency in IUGR fetuses. In fact, in our IUGR group of fetuses, we found the presence of hemodynamic changes essentially characterized by brain sparing in the middle cerebral artery, without signs of hemodynamic decompensation17-19. The relative greater blood supply, in the cerebral area, due to the reactive vasodilatation, would be required to maintain oxygen supply and enhanced growth.

Possibly, this could also be related to the edema of these cerebral structures, due to the increase of vascular permeability related to the vasodilation and to the release of pro-inflammatory vasoactive substances in response to hypoxia19,22. Realistically, the edema gradually decreased after birth, due to the disappearance of the oxygen deficiency and brain sparing. This process may be related with the reduced volume of these brain structures as observed in literature of post-natal studies23. The observed differences in the cerebral volume of IUGR cases, despite being encouraging, should be interpreted with caution and further studies with a larger population are necessary to confirm these findings.

Diabetes mellitus occurs in about 7% of pregnant women24 and represents the most serious metabolic alteration in pregnancy, with a significant increased risk of maternal-fetal complications25. The fetal brain is particularly susceptible to the maternal hyperglycemia26. Maternal hyperglycemia, during pregnancy, could act through three major mechanisms: hypoxia, oxidative stress and increased inflammation27,28. The consequences of these mechanisms of action at various levels can lead to permanent effects on fetal cellular physiology associated with a spectrum of cellular anomalies and even with schizophrenia28. The same mechanisms are responsible for alterations in fetal growth or fetal development. Currently, existing studies on this field were conducted by RMN and the brain volumes were calculated on diabetic adult subjects and not on the fetus in utero29. In a recent study, Hami et al analyzed the effect of maternal diabetes and insulin treatment on cerebellar volume and morphogenesis of the rat cerebellar cortex in the first two weeks after birth29. The authors concluded that diabetes in pregnancy interferes with cerebellar morphogenesis in rat. In the cerebellum of maternal diabetic neonatal rats, the density of Granular cells and Purkinje cells was increased and their number progressively decreased after birth. In our study the thalamic volumes resulted significantly bigger in diabetic group than control, while for cerebellum and unilateral cerebrum absolute values were higher in diabetes but these differences did not achieve the statistical significance probably in relation to the small number of cases (6 cases of cerebellar volumes in 15 diabetic women). Examining the other variables: “fetal weight at birth”, “APGAR score” and “centile of growth”, we did not find a significant difference probably in relation to the appropriate management of patients with gestational diabetes. Finally, the variable “gestational age at birth” was significantly lower in diabetic group (p = 0.01) most likely related to

the use of labor induction in this group. Further studies with a larger population are necessary to confirm these results.

In conclusion, our study suggests that it’s possible to obtain fetal cerebral structures volumetric growth curves by 3D ultrasound examination investigating fetuses between 17 and 39 weeks of gestation. The use of tissue volumes for the construction of growth curves could add important information to the biometric 2D curve. Furthermore, this study suggests that the fetal brain is influenced in different ways, when exposed to a specific insult (like oxygen deficiency in IUGR and hyperglycemia in diabetes). For those reasons the fetal brain can be considered a dynamic structure that changes its response depending on the type, amount and characteristics of the injury. For the future it would be interesting to continue sampling the thalamic, unilateral cerebrum and cerebellar volumes by increasing IUGR and diabetic cases to validate the hypotheses described above.

Figure 1A

Figure 1B

Figure 2A

Figure 2B

Figure 3 A

Figure 3B

FIGURE 4

FIGURE 5

FIGURE 6

FIGURE 7

GC has created the idea of the study and the main modalities of development, he has also partially sampled the fetal brain volumes

GB partially sampled the fetal cerebral volumes and recalculated all volumes on them with specific applications

BC contributed to the pediatric part of the study with controls on newborns

GCdR reviewed the study design and provided data analysis

KGR contributed to the improvement of the language of writing the manuscript and carried out a critical judgment of the work

Figure 1A

Multi-planar view of the thalamus: axial (as working plane), coronal and sagittal views

Figure1B

Rendering of thalamic volume expressed in cm3

Figure 2A

Multi-planar view of the cerebellum: axial (as working plane), coronal and sagittal views

Figure2B

Rendering of cerebellum volume expressed in cm3

Figure 3A

Multi-planar view of the hemi-cortex: axial (as working plane), coronal and sagittal views

Figure3B

Rendering of hemi-cortex volume expressed in cm3

Figure 4

Thalamic volume growth curve (class 1, 21+0; class 2, 21+1– 21+2; class 3, 21+3- 21+6; class 4, 21+6-22+3; class 5, 22+4-25+5; class 6, 25+6-27+0; class 7, 27+1-29+6; class 8, 30+0-33+1; class 9, 33+2-38+5)

Figure 5

Cerebellum volume growth curve (class 1, 21+0; class 2, 21+1– 21+2; class 3, 21+3- 21+6; class 4, 21+6-22+3; class 5, 22+4-25+5; class 6, 25+6-27+0; class 7, 27+1-29+6; class 8, 30+0-33+1)

Figure 6

Hemi-cortex volume growth curve (class 1, 21+0; class 2, 21+1– 21+2; class 3, 21+3- 21+6; class 4, 21+6-22+3; class 5, 22+4-25+5; class 6, 25+6-27+0, class 7, 27+1-29+6; class 8, 30+0-33+1; class 9, 33+2-38+5)

Figure 7

Cortex volume growth curve (class 1, 21+0; class 2, 21+1– 21+2; class 3, 21+3- 21+6; class 4, 21+6-22+3; class 5, 22+4-25+5; class 6, 25+6-27+0, class 7, 27+1-29+6; class 8, 30+0-33+1; class 9, 33+2-38+5)

References

1. Correa FF, Lara C, Bellver J, Remohi J, Pellicer A, Serra V. Examination of the fetal brain by transabdominal three-dimensional ultrasound: Potential for routine neurosonographic studies. Ultrasound Obstet Gynecol 2006;27:503–508.

2. Tonni G, Martins WP, Filho HG, Araujo EJ. Role of ultrasound in clinical obstetric practice: evolution over 20 years. Ultrasound in Med & Biol 2015 1-32

3. Endres LK1, Cohen L. Reliability and validity of three-dimensional fetal brain volumes J Ultrasound Med. 2001 Dec;20(12):1265-9.

4. Roelfsema NM, Hop WC, Boito SM, Wladmiroff JW. Three-dimensional sonographyc measurement of normal fetal brain volume during the second half pregnancy. AM J Obstet Gynecol. 2004Jan;190(1):275-80.

5. Caetano AC, Zamarian AC, Araujo Júnior E, Cavalcante RO, Simioni C, Silva CP, Rolo LC, Moron AF, Nardozza LM. Assesment of Intrcranial structure volumes in fetuses with growth restriction by 3-dimensional sonography using the extended imaging Virtual Organ Computer-Aided Analysis method. J ultrasound Med 2015; 34:1397-1405.

6. Pavlova E, Markov D, Ivanov S. Three-dimensional (3D) ultrasound in evaluation of fetal brain anatomy. Akush Ginekol (Sofiia). 2014;53(2):11-7.

7. Pooh RK, Pooh K, Nakagawa Y, Nishida S, Ohno Y . Clinical application of three-dimensional ultrasound in fetal brain assessment. Croat Med J. 2000 Sep;41(3):245-51.

8. Allin MP1. Novel insights from quantitative imaging of the developing cerebellum. Semin Fetal Neonatal Med. 2016 Oct;21(5):333-8. 2016 Jun 26.

9. Shmahmann JD. Disorders of the cerebellum: ataxia, dysmetria of thought, and the cerebellar cognitive affective syndrome. J Neuropsychiatry Clin Neurosci 2004;16:367-78.

10. Sotiriadis A1, Dimitrakopoulos I, Eleftheriades M, Agorastos T, Makrydimas G. Thalamic volume measurement in normal fetuses using three-dimensional sonography. J Clin Ultrasound. 2012 May;40(4):207-13.

11. Tamura R, Kitamura H, Endo T, et al. Reduced thalamic volume observed across different subgroups of autism spectrum disorders. Psychiatry Res 2010;184:186.

12. Miller SL, Petra S, Mallard HC. The consequences of fetal growth restriction on brain structure and neurodevelopmental outcome. J Physiol 2016 pp807-23.

13. Chang CH1, Chang FM, Yu CH, Ko HC, Chen HY. Assessment of fetal cerebellar volume using three-dimensional ultrasound. Ultrasound Med Biol. 2000 Jul;26(6):981-8.

14. Araujo Júnior E, Pires CR, Nardozza LM, Filho HA, Moron AF Correlation of the fetal cerebellar volume with other fetal growth indices by three-dimensional ultrasound. J Matern Fetal Neonatal Med. 2007 Aug;20(8):581-7

15. Kyriakopoulou V, Vatansever D, Davidson A, Patkee P, Elkommos S, Chew A, Martinez-Biarge M, Hagberg B, Damodaram M, Allsop J, Fox M, Hajnal JV, Rutherford MA. Normative biometry of the fetal brain using magnetic resonance imaging. Brain Struct Funct. 2016 Nov 24.

16. Benavides-Serralde A, Hernàndez-Andrade E, Fernàndez-Delgado J, Plasencia W, Scheier M, Crispi F, Figueras F, Nicolaides KH, Gratàcos E. Three-dimensional sonographic calculation of the volume of intracranial structures in growth-restricted and appropriate-for-gestational age foetuses. Ultrasound Obstet Gynecol. 2009 May;33(5):530-7

17. Tolsa CB, Zimine S, Warfield SK, Freschi M, Lazeyras F, Hanquinet S, Pfizenmaier M, Huppi PS. Early alteration and functional brain development in premature infants born with intrauterine growthrestriction. Pediatr Res 2004; 56:132-8.

18. Gramsbergen A. Neural compensation after early lesions: a clinical view of animal experiments. Neurosci Biobehav Rev 2007; 31:1088-94.

19. Hernàndez-Andrade E, Figueroa-Diesel H, Jansson T, Rangel-Nava H, Gratacos E. Changes in regional fetal cerebral bloodflow perfusion in relation to hemodynamic deterioration inseverely growth-restricted fetuses. Ultrasound Obstet Gynecol 2008; 32: 71-6.

20. Yang C1, Liu Z, Li H, Zhai F, Liu J, Bian J. Aquaporin-4 knockdown ameliorates hypoxic-ischemic cerebral edema in newborn piglets. IUBMB Life. 2015 Mar;67(3):182-90.

21. Goyal R1, Van Wickle J2, Goyal D2, Matei N2, Longo LD1. Antenatal maternal long-term hypoxia: acclimatization responses with altered gene expression in ovine fetal carotid arteries. PLoS One. 2013 Dec 18;8(12):e82200J.

22. Xu Y1, Zuo Y, Zhang H, Kang X, Yue F, Yi Z, Liu M, Yeh ET, Chen G, Cheng J. Induction of SENP1 in endothelial cells contributes to hypoxia-driven VEGF expression and angiogenesis. J Biol Chem. 2010 Nov 19;285(47):36682-8

23. Rogne T1, Engstrøm AA, Jacobsen GW, Skranes J, Østgård HF, Martinussen M. Fetal growth, cognitive function, and brain volumes in childhood and adolescence. Obstet Gynecol. 2015 Mar;125(3):673-82

24. Reece EA, Homko CJ. Diabetes mellitus in pregnancy. What are the best treatment options? Drug Saf 1998; 18:209-210

25. Metzger BE. Summary recommendations of the third international workshop-Conference on Gestational Diabetes Mellitus Diabetes 1991;40(suppl2):197-201

26. Cannon M, Jones PB, Murray RM. Obstetric complications and schizophrenia: historical and meta-analytic review. Am J Psychiatry 2002;159:1080-92.

27. Shwartz MV, Porte D Jr. Diabetes, obesity and the brain. Science 2005; 307:375-9.28. Ryan J, Van Lieshout and Lakshmi P, Vourganti. Diabetes mellitus during pregnancy and increased

risk of schizophrenia in offspring: a review of the evidence and putative mechanisms. J Psychiatry Neurosci. 2008 sep; 33(5):395-404.

29. HamiJ, Vafaei-Nezhad S, Ghaemi K, Sadeghi A, Ivar G. Shojae F, Hosseini M. Stereological study of the effects of maternal diabetes on cerebellar cortex development in rat. Metab Brain Dis. 2016 Jun; 31 (3):643-52.