Recent Results and Future Challenges in the Assessment of Fetal Brain Function

Structural Development of CNS

The development of CNS begins approximately at the end of gastrulation. The generation of the neuroectoderm from ectoderm during the third postconceptional week results in the formation of the neural plate. Thus, the neural epithelium of the embryo, which is a precursor of neurons and glia, is virtually the first part of an organism that acquires a separate identity from other cells.² The formation of the neural plate is succeeded by the folding of its edges and formation of a neural tube, whose further growth and reshaping results in the formation of structures of CNS. According to O’Rahilly and Muller, the forebrain (prosencephalon), midbrain (mesencephalon), and hindbrain (rhombencephalon) can be distinguished in the rostral portion of the unfused neural folds³ earlier than it is usually referred to, approximately at the 22nd postconceptional day. In the rapid succession, during the 4th postconceptional week, the forebrain components—diencephalon and telencephalon can be detected. Three embryonic zones, ventricular, intermediary, and marginal zone (seen from ventricular to the pial surface), are present in all parts of the neural tube, while telencephalon contains additional two zones, subventricular and subplate zone, Ventricular zone and subventricular zone of telencephalon are the site of neurogenesis and all the future neurons and glia are born in these structures. During migration toward the pial surface, they form other transitional zones before reaching their genetically predetermined final destinations. Those destinations are cortical plate or different nuclei in the brain stem, diencephalon, and basal forebrain [for references, see¹]. One of the transitional structures, a subplate zone that is a site for transient synapses and neuronal interactions, can play a major role in the developmental plasticity following perinatal brain damage.⁴ Early appearance of interneuronal connections, shown in Table 1, implicates a possibility of early functional development. However, these first synapses exist only temporarily and disappear due to the normal reorganization processes. Most embryonic zones, types of neurons and glia, and early synapses, which play a crucial role in certain periods of fetal brain development, eventually disappear, significantly changing the structure and function of the brain. Reorganization processes include apoptosis, the disappearance of redundant synapses, axonal retraction and transposition, and the transformation of the neurotransmitters phenotype.⁴

Table 1 shows a significant overlap of neurogenesis, migration, and synaptogenesis in embryonic and fetal life. The development of the human brain is not completed at the time of delivery. In an infant born at term, characteristic cellular layers can be observed in motor, somatosensory, visual, and auditory cortical areas. Although proliferation and migration are completed in a term infant, synaptogenesis and neuronal differentiation continue very intensively.⁵ Brain stem demonstrates a high level of maturity, whereas all histogenetic processes actively persist in the cerebellum.⁶ Therefore, only subcortical formations and the primary cortical areas are well developed in a newborn. The associative cortex, barely visible in a newborn, is scantily developed in a 6-month-old infant. Postnatal formation of synapses in associative cortical areas, which intensifies between the 8th month and 2nd year of life, precedes the onset of first cognitive functions, such as speech. Following the 2nd year of age, many redundant synapses are eliminated. The elimination of synapses begins very rapidly and continues slowly until puberty when the same number of synapses as seen in adults is reached.⁶

Functional Development of CNS and Role of Four-dimensional Ultrasound (4D US)

The first synapses appear in the spinal cord at 6–7 postconceptional weeks⁷ and in the cortical plate at 8 postconceptional weeks.⁸ This is the phase when the first electrical bustle and conduction of information take place. The earliest spontaneous fetal movements can be observed at 7.5 postconceptional weeks. These movements, consisting of slow flexion and extension of the fetal trunk, accompanied by the inactive displacement of arms and legs and emerging in asymmetrical sequences, have been described as “vermicular”.^9,10 They are substituted by various general movements (GMs), which consist of head, trunk, and limb movements, such as “rippling” seen at week 8, “twitching” and “strong twitching” at weeks 9 and 9.5, respectively, and “floating” “swimming” and “jumping” at week 10.¹¹ Isolated limb movements emerge almost simultaneously with the GMs. At the same time with the beginning of spontaneous movements, at 7.5 postconceptional weeks, the initial motor reflex activity can be detected, permitting the hypothesis to be made of the existence of the first afferent–efferent circuits.⁷ At that time, head tilting following perioral stimulation was noted. The primary reflex movements are immense and signify a limited number of synapses in a reflex pathway. During the 8th week of gestation, these substantial reflex movements are replaced with local movements, possibly due to an increase in the number of axodendritic synapses. Hands become susceptible at 10.5 weeks and lower limbs start to contribute to these reflexes at around week 14. General movements are the first sign of a supraspinal control on fetal motor activity. Approximately at the 7 postconceptional weeks, the brainstem which consists of the medulla oblongata, pons, and midbrain, begins to develop and mature in a caudal to the rostral direction. As the medulla matures in advance of more rostral structures of the brainstem, reflexive movements of the head, body, and extremities, as well as breathing movements and heart rate alterations, appear in advance of other functions.¹² From the 10th week onward, the amount and incidence of movements increase. By 14–19 weeks, fetuses are highly active with the longest period between movements of only 5–6 minutes. In the 15th week, 15 singular types of movement can be observed. The general body movements and isolated limb movements, retroflexion, anteflexion, and rotation of the head can be seen. Also, face movements, such as mouthing, yawning, hiccups, sucking, and swallowing, can be included to an ample repertoire of fetal motor activity in this stage.¹³ But, during the first half of pregnancy, a dynamic pattern of neuronal production and migration, as well as the immature cerebral circuits are considered too immature for cerebral involvement in the motor behavior.⁴ Merely at the end of this period do a quantifiable number of synapses appear in the structures preceding the cerebral cortex, perhaps forming a substrate for the first cortical electric activity, noted at week 19.⁴ The spinothalamic tract is established at the 20th week and myelinized by 29 weeks of gestation and the thalamocortical connections penetrate the cortical plate at 24–26 weeks. Evoked potentials can be detected from the cortex at the 29th week, indicating that the functional connection between periphery and cortex operates from that time onward.¹⁴ In the second half of pregnancy, the number of GMs gradually decreases, particularly during the last 10 weeks.¹⁵ Although this decrease was first explained as a result of the reduction in amniotic fluid volume, it is now believed to be a result of maturation processes in the brainstem.¹² Simultaneously with the decline in the number of generalized movements, an increase in facial movements, as well as opening or closing of the jaw, swallowing, and chewing, can be observed. These activities can be seen mainly in the periods of absence of GMs, and this pattern is considered to be a manifestation of the normal neurological development of the fetus.¹⁵ However, alterations not only in the number of movements but also in their complexity, are revealed to be the result of cerebral maturation processes. It is important to point out that subunits of the brainstem remain the main regulators of all fetal behavioral patterns until delivery.¹² Despite medical reports from 100 years ago and >30 years of systematic research initiated by Prechtl and colleagues,^16–18 the study of prenatal behavior is still in its infancy. One of the most promising advances in the field of ultrasonography has been the new 4D US technology. Its advance has been completed in the last year giving visualizations in almost real-time.^19–22 The availability of new diagnostic data has in an extraordinary way raised our knowledge about intrauterine life, substantially modifying some earlier interpretations.²³ First spontaneous fetal movements can be observed with conventional two-dimensional ultrasound (2D US) around 8th gestational week, while the newly developed 4D US allows the visualization of fetal motility 1 week earlier (Table 2).²⁴

General movements are the first complex fetal movement patterns observable by two-dimensional (2D) US, but assessment by the 4D US is a significant improvement. They can be recognized from the 8th to 9th weeks of pregnancy (Fig. 1 showed by 4D US) and continue to be present until 16–20 weeks after birth.¹⁸

According to Prechtl, these are gross movements, involving the whole body. They wax and wane in intensity, force, and speed, and they have a gradual beginning and end.^13,18 The majority of sequences of extension and flexion of the legs and arms is complex and may be better assessed with 4D US.²⁴ In the literature, there is a range between the 8th and 12th weeks concerning the first appearance of limb movements.^13,18,20,25 Isolated arm and leg movements found by de Vries et al. at the 8th week of gestation.¹³ With 4D US, limb movements at the 8th to 9th weeks were found. Using 4D sonography, Kurjak et al. found that from 13 gestational weeks onward, a “goal orientation” of hand movements appears and a target point can be recognized for each hand movement.¹⁹ More limb joints are active and move simultaneously, such as extension or flexion in arm and elbow or hip and knee. The elevation of the hand, extension of the elbow joint, with a slight change in direction and rotation, can be seen simultaneously.²⁶ The isolated limb movements seen at the 9th week are followed by the appearance of the movements in the elbow joint at 10 weeks, changes in finger position in the 11th week and by easily recognizable clenching and unclenching of the fist at the 12th to 13th weeks. Finally, at the 13th to 14th weeks, isolated finger movements can be seen, as well as an increase in the activity and strength of the hand or finger movements.²⁶ Recent investigation of fetuses in the last trimester of gestation, performed by 4D US, has discovered an even wider range of hand and face movements than was formerly explained.¹⁹ It has been also confirmed that the fetal movement patterns in the second half of pregnancy are about equal to those monitored after birth, while the list of movements in the newborn consists of some patterns that cannot be watched in the fetus, such as the Moro reflex.²⁷ As well, study of anencephalic fetuses has presented clear evidence for the influence of supraspinal structures on motor behavior at around the 20th gestational week. In these fetuses, the number of movements was normal or even increased, but the complexity of the movement patterns distorted radically and movements were stereotyped and simplified.²⁸

The eminence of fetal movement patterns is distorted in fetuses undergoing intrauterine growth restriction. The activities become slower and monotonous, similar to cramps and their variability in force and amplitude is reduced.²⁹ These changes might designate the subsistence of brain lesions in growth-restricted and possibly hypoxic fetuses. Despite the premature postulations, the changes in the amplitude and complexity of movements in these fetuses do not show to be due to oligohydramnios. In cases of premature rupture of fetal membranes and a subsequently reduced volume of amniotic fluid, movements arise less frequently, but their complexity look likes that of movements achieved in the normal volume of amniotic fluid.¹⁶ Qualitative including quantitative analysis of fetal movements divulged the consistency of the fetal nervous system, and can be applied for the recognition of different cerebral dysfunctions, and probably neuromuscular ailments.

The application of the new technology, 4D US, in the examination of fetal facial movements has revealed the existence of a full range of facial expressions including grimacing, tongue expulsion, and eye-lid movements (Fig. 2) similar to emotional expressions in adults.^27,30 The possibility of studying such subtle movements might open a new area of investigation.³¹

During the first trimester, a tendency toward increased frequency of fetal GMs with increasing gestational age has been noticed (Fig. 3A). While at the beginning of the second trimester, the fetuses began to display a tendency toward increased frequency of observed fetal facial expression up to the end of the second trimester. An oscillation and dispersion of the incidence of the facial expression as seen in the polynomial regression of isolated eye blinking diagram is observable in Figure 3B.³⁰

The most frequent facial movement patterns in the second trimester were isolated eye blinking, grimacing, suckling, and swallowing while yawning, mouthing, tongue expulsion, and smiling could be observed less frequently.^25,30 During the third trimester, the fetuses began to display decreasing or stagnant incidence of fetal facial expression. All of the types of head movements and hand-to-body contact indicated a tendency to decrease frequency from the beginning of the second trimester to the end of the third trimester (Fig. 3C).³⁰

The investigations of fetal facial expressions established that all mechanisms of the fetal yawning pattern, prolonged jaw opening followed by a quick closure and accompanied by head flexion and elevation of arms, can easily be documented by 4D US in this period (Fig. 4).³²

If the fetal yawning in the third trimester was matched up to the yawning in the neonates during the first week of life, no differences were found in the frequencies of this reaction. The frequency of yawning steadily increased between the 15th and 24th week when a short plateau was observed from 24th to 26th week and was followed by a slight decrease toward the term.³⁰ A gestational age-related tendency in the frequency of yawning could be assumed as the maturation of the brain stem and probably the gaining of control of more cranial structures over yawning pattern. These results have offered new data concerning the route of neurodevelopment of this fascinating, but poorly implicit reflex. Whether this is distorted in cases of neurodevelopmental disorders and whether such adaptations can give us impending into the function of fetal nervous system in high-risk pregnancies, continues to be determined. It also stays to be determined to what possibility are the facial motoric patterns related to the function and integrity of the CNS. However, the fact that even in the embryonic stage, the identical inductive forces that cause growth and restyling of the neural tube influence the development of facial structures, and that many genetic disorders affecting the CNS are also described by dysmorphology and dysfunction of facial structures, underline the value of these studies.^2,30,31

Neonatal Aspect of Fetal Neurology

Obstetric aspects of fetal neurology have been discussed extensively in our previous publications.^{19–25,27–31} To better understand fetal neurobehavioral patterns, we have learned a lot from basic studies of brain development and clinical postnatal studies of neonates. Now, when we have reached the edge of a fetal behavioral investigation by 4D US in normal fetuses, we intend to find some new ideas and ways of investigation presenting a neonatal aspect of fetal neurology.

Cerebral palsy is an “umbrella” term for disorders of development, movement, and posture, resulting in limitations of activity due to nonprogressive impairment of the developing brain.³³ The diagnosis of CP is retrospective and it is exceptionally made before the age of 6 months in only the most severely affected infants, and the specificity of the diagnosis will improve as the child ages and the nature of the disability evolves.³⁴ Cerebral palsy does not result from a single event but rather there is a sequence of interdependent adverse events providing to the condition.³⁵ This time frame of evolving adverse events is something that should be taken into account when considering the possibility of CP diagnosis in infants.^34,35 The understanding of the profile of a child’s disability across multiple domains is an ongoing process necessary for appropriate treatment and future planning.³⁴ This theoretical statement is sometimes very difficult to be practically implemented. An attempt to make an early diagnosis of CP should be followed with factors related to pathogenesis, impairment, and functional limitations in every patient.³⁴ To identify the pathogenesis of the process, neuroimaging methods should be used, among which cranial ultrasound, magnetic resonance imaging (MRI), magnetic resonance spectroscopy, and diffusion-weighted imaging are the most frequently used in very low birth-weight premature infants and term infants with encephalopathy.³⁴ Impairment of organs or systems by clinical assessment of muscle tone, strength, and control of voluntary movements for early detection of infants with the risk for CP has been frustrating because only 43% of 7-year-old children with CP had a normal newborn neurological examination.^34,36 Is it possible to change this discouraging fact resulting from our failure to diagnose neurological impairment early enough to intervene? Interests in the diagnosis of neurological impairment among ultrasonographers using 4D US have been recently shifted toward the prenatal period.^23,37 Is there any possibility to improve the timing of the postnatal diagnosis of a neurologically disabled infant? Postnatal assessment is probably easier to perform than prenatal, by using a simple and suitable for everyday work screening clinical test with good reliability, specificity, and sensitivity. Such tests are still not widely used, while those complicated and time-consuming are used mostly for clinical research purposes. There is a possibility for the early and simple neurological assessment of the term and preterm newborns to detect associated risks and anticipate the long-term outcome of the infant, and to establish a possible causative link between pregnancy course and neurodevelopmental outcome.³⁸ As CP is a disorder of movement and postural control resulting in functional limitations, its diagnosis could be helpful in the detection of early impairment.³⁴ Clinical neurological assessment proposed and practiced by Amiel-Tison could be very useful in the early detection of newborns at risk.³⁸ As the development of the CNS is a very complex and long-lasting process, the assessment of its developmental optimality is something that should be assessed to investigate whether the infant is neurologically normal or damaged. Neurological assessment at term by Amiel-Tison is taking into account neurological maturation exploring so-called lower subcortical system developing earlier from the reticular formation, vestibular nuclei and tectum, and upper cortical system developing from the corticospinal pathways.³⁹ The role of the lower system is to maintain posture against gravity, while the upper system is responsible for the control of erect posture and the movements of the extremities.³⁹ At the corrected age of 40 gestational weeks optimality assessment consists of head circumference measurement, assessment of cranial sutures, visual pursuit, social interaction, sucking reflex, raise-to-sit and reverse, the passive tone in the axis, the passive tone in the limbs, fingers, and thumbs outside the fist, and autonomic control during the assessment.³⁹ The Amiel-Tisson neurologic assessment at term is increasing accuracy in assessing CNS function in the neonate by using a simple scoring system, focusing on the most meaningful items, promoting a clinical synthesis at term, for the term and preterm infants.³⁹ It was recognized that clinico-anatomic correlations using high-resolution neuroimaging techniques could be helpful in the neurological assessment of newborns, while the neurological examination and the functional assessment of the developing CNS are bringing a new perspective of CNS status in the neonatal period.⁴⁰

Postnatal Assessment of GMs

In the last 30 years, objective assessment of videotaped GMs by Prechtl’s method is predictive of later CP.^9,16,18,41 The quality of GMs at 2–4 months post-term (so-called fidgety GM age) has been found to have the highest predictive value in the detection of infants at risk for CP development.⁴² It seems that assessment of the quality of GM is a window for early detection of children at high risk for developmental disorders.^17,42 The method is simple and it is based on the so-called Gestalt evaluation of GM complexity and variation.^17,41,42 Assessment of GMs at 2–4 months post-term at so-called fidgety GM age has been found to have the highest predictive value for the development of CP if abnormal.^17,41,42

Heinz Prechtl’s work enabled that spontaneous motility during human development has been brought into the focus of interest of many perinatologists prenatally and developmental neurologists postnatally.^{9,16,18,41,42} According to the research preceding Prechtl’s ingenious idea, during the development of the individual the functional repertoire of the developing neural structure must meet the requirements of the organism and its environment.⁴¹ This concept of ontogenetic adaptation fits excellently to the development of the human organism, which is during each developmental stage adapted to the internal and external requirements.⁴¹ Prechtl stated that spontaneous motility, as the expression of spontaneous neural activity, is a marker of brain proper or disturbed function.^41,42 The observation of an un-stimulated fetus or infant which is the result of spontaneous behavior without sensory stimulation is the best method to assess its CNS capacity.⁴¹ All endogenously generated movement patterns from the un-stimulated CNS could be observed as early as from the 7–8 weeks of postmenstrual age, with developing a reach repertoire of movements within the next 2 or 3 weeks, continuing to be present for 5–6 months postnatally.¹³ This remarkable fact of the continuity of endogenously generated activity from prenatal to postnatal life is a great opportunity to find out those high-risk fetuses and infants in whom the development of neurological impairment is emerging. The most important among those movements are GMs involving the whole body in a variable sequence of arm, leg, neck, and trunk movements, with gradual beginning and the end. They wax and wane in intensity, force, and speed being fluent and elegant with the impression of complexity and variability. General movements are called fetal or preterm from 28 to 38 weeks of postmenstrual age, while after that we have at least two types of movements: writhing present to 46 to 52 weeks of postmenstrual age and fidgety movements present till 54 to 58 weeks of postmenstrual age.^18,41,42 Lack of fluency and the existence of considerable variation and complexity are the main characteristics of mildly abnormal GMs.⁴³ When complexity, variation, and fluency are absent then we are dealing with definitely abnormal GMs.⁴³

The quality of each individual movement includes speed, amplitude, and force combined in one complex perception.^{13,18,41–44} Investigation of normal and neurologically impaired preterm infants showed that except for the higher incidence of clonus in the abnormal group, there was no marked difference in the number of different motor patterns studied.^44,45 However, video analysis of another group of sick preterm infants revealed a “reduction of elegance and fluency as well as variability, fluctuation in intensity and speed rather than any change in the incidence of distinct motor patterns”.^44–46 Based on postnatal studies, it would be very important to seek for abnormal quantity and quality of prenatal movements to find fetuses neurologically at risk.⁴⁶

Some facts are very important in the assessment of GMs. The first is that the evaluation of GMs should be based on the video recorded movements either pre- or postnatally. The second fact is that when assessing GMs one should use so-called “gestalt perception”, which could be described as the overall impression of GMs with a standardized procedure.⁴¹ During the perception, one should recognize the movement patterns of GMs, then assess their complexity, variability, and fluency.^41,42 According to Hadders-Algra, GMs could be classified as normal-optimal, normal-suboptimal, mildly abnormal, and definitely abnormal.⁴² This modality of GM assessment is important for the prenatal and postnatal observation of GMs. It is not so important to assess the quantity of GMs, while the assessment of their quality is of utmost importance in terms of the prognosis of the neurodevelopmental outcome. They can better predict neurodevelopmental outcomes than classical neurologic examination alone.⁴⁷

We can conclude that prenatal and postnatal assessment of GMs according to Prechtl’s method gives quite new insight into the function and development of CNS. This important modality is time-consuming and requires some technology and expertise to be practiced, but the advantages of its implementation in prenatal and postnatal life are very promising and encouraging in terms of its prognostic value. Prenatal assessment of GMs is well developed and established, while prenatal assessment needs sophisticated real-time 4D ultrasonographic or other technology to enable a more precise assessment of GM quality in fetuses.

Continuity of GMs from Prenatal to Postnatal Life

Postnatal studies of neonatal behavior have taught us that the assessment of behavior is a better predictor of neurodevelopment disability than neurological examinations.⁴⁶ It is important to mention that postnatal observation of movement patterns was introduced by Prechtl and coworkers in the way that they have been observing spontaneous movements of the infant using video typing and “off-line” analysis of both quantity and quality of the movement.^17,48 They proved that assessment of GMs in high-risk newborns has significantly higher predictive value for later neurological development than neurological examination.^46,47,49 Kurjak and coworkers conducted a study by 4D US and confirmed earlier findings made by the 2D US, that there is behavioral pattern continuity from prenatal to postnatal life.²⁷ Assessment of neonatal behavior is a better method for early detection of CP than neurological examination alone.⁵⁰ It is being speculated that intrauterine detection of encephalopathy would improve the outcome. Although many fetal behavioral studies have been conducted, it is still questionable whether the assessment of continuity from fetal to neonatal behavior could improve our ability of early detection of brain pathology. Early detection could possibly rise an opportunity to intervene and even prevent the expected damage.

Could Some Postnatal Signs of Neurological Disability be Used Prenatally?

It has been proven by now that ultrasonography is a powerful tool in the assessment of fetal behavior. Four-dimensional ultrasound brought up to light visual observation of the fetus, particularly in two especially important domains: fetal finger movements and facial expressions.^19,51 This new technology is not only a toll of fetal observation but a very useful toll to evaluate the development of fetal CNS in normally developing fetuses and those at high risk. A basic understanding of fetal neurology include: defining of motor pathways involved, the chronology of their maturation, and the direction of myelination.^52,53 This information helps the clinician in better interpretation of fetal movements. The experience acquired with Amiel-Tison’s Neurological Assessment at Term (ATNAT) helps us in the interpretation of fetal movements.^39,54,55

The domain of fetal neurology is already too extensive, but the focus of interest is mainly the second trimester, even though spontaneous fetal mobility emerges and has already became differentiated at a very early age.⁵⁶ This means that we will take into consideration the period of pregnancy from 20 till 40 weeks of gestation, including the end of the neuronal migration and the post-migratory phase corresponding to the development of the neocortex.^4,57

As it was already mentioned, CP describes a group of disorders of the development of movement and posture, causing activity limitations, which are attributed to nonprogressive disturbances occurring at the time of fetal brain development.^58–61 Motor disorders that occur in patients with CP are often accompanied by disturbances of sensation, cognition, communication, perception, behavior, and/or seizure disorder.^58–61 “Disturbances” is a term that refers to events or processes that in some way influence the expected pattern of brain maturation.⁵⁵ Those events or processes are many, with consequences varying from very conspicuous to very subtle. It should be kept in mind what many neurologists emphasize, that morphology does not always correspond to the neurological outcome.^39,54,55 The opposite view is the one from pediatricians and neurophysiologists, who are involved in long-term follow-up studies, and they are certainly not that optimistic. It would be wise to consider the long-run prognosis, for each specific type of fetal brain damage and make appropriate decisions for conservative management.

Hopes have been headed toward magnetic resonance (MR), but in many cases brain changes cannot be detected as early as the first year of life, like pathological gliosis which causes secondary hypomyelinization.

While examining the fetal head by 4D, the sonographer should examine bony structures and fetal cranial sutures, if they are folding over one another, it is considered to be a bed sign previously described by Amiel-Tison.^39,55 The same sign should be searched for postnatally, as a part of the neurological examination.⁶²

The majority of pediatricians believe that the main obstacle for early prediction of CP based on a functional observation of the fetus such as visual observation by the 4D US is due to the “precompetent” stage of most of the motor behavior observed in utero.^39,55 One of the possible signs detected could be a high arched palate, described by Amiel-Tison, in the clinical assessment of the infant nervous system.^39,55 What was believed as undetectable became visible by 4D. Recently, the 3D “reverse face technique” has been described. This technique overcomes shadowing the fetal face by rotating the frontal facial image through 180° along the vertical axis, so that the palate, nasal cavity, and orbits become visualized.^63,64

In their early work, Pooh and Ogura examined 65 normal fetuses by 3D/4D. The purpose of their study was to investigate the natural course of fetal hand and finger positioning.²⁶ During the 9th and at the beginning of the 10th week fetal hands were located in front of the chest and no movements of wrists and fingers were visualized. From the middle of the 10th week, active arm movements were observed.²⁶ This study is very important because it is showing that finger and thumbs movements begun in the early stage of human life, long before the maturation of the upper system. Therefore, this motor activity depends on the lower system and not before 30 and 32 weeks switches to the upper control.

Amiel-Tison also described the so-called neurologic thumb squeezed in a fist. Clenched fingers can also be detected by the 4D US, as well as overlapping cerebral sutures.^19,26

Head anteflexion becomes visible during the 10th and 11th gestational week, according to de Vries et al.⁵⁶ However, the activity of flexor muscles will depend on the upper system since 34 weeks of gestation. The absence of active head flexion explored by the raise-to-sit maneuver is one of the major neurological signs at 40 weeks of gestation.^39,54,62

What have Studies about KANET Shown So Far?

One of the first studies to use a preliminary form of the KANET scoring system was that by Andonotopo and Kurjak in 2006.²⁹ They aimed to assess fetal facial expression and quality of body movements and examine if they are of diagnostic value for brain impairment in fetuses with growth restriction. In that prospective study of 50 pregnancies with IUGR fetuses in the 3rd trimester of pregnancy, a tendency of less behavioral activity in IUGR than normal fetuses has been noted. The results of the study encouraged the future investigation of the use of 4D US for quantitative and qualitative assessment of fetal behavior as possible indicators of the neurological condition in IUGR fetuses.²⁹

The Zagreb group in 2008 was the first to introduce the KANET for the assessment of the neurological status of the fetus, aiming to the detection of the fetal brain and neurodevelopmental alterations due to in utero brain impairment.²⁷ To develop the new scoring system they identified severely brain-damaged neonates and neonates with a good neurological condition and then compared the neonatal findings, with corresponding findings in utero. In the group of 100 low-risk pregnancies, they retrospectively applied KANET. After delivery, postnatal neurological assessment (ATNAT) was performed and all neonates assessed as normal reached a score between 14 and 20, assumed to be the score of optimal neurological development. A new scoring system was applied in the group of 120 high-risk pregnancies in which, based on postnatal neurological findings, three subgroups of newborns were identified: normal, mildly or moderately abnormal, and abnormal.³⁰ Based on this, a neurological scoring system has been proposed. All normal fetuses reached a score from 14 to 20. Ten fetuses who were postnatally described as mildly or moderately abnormal achieved a prenatal score of 5–13, while another 10 fetuses postnatally assigned as neurologically abnormal had a prenatal score of 0–5. Among this, group 4 had alobar holoprosencephaly, one had severe hypertensive hydrocephaly, one had thanatotrophic dysplasia, and four fetuses had multiple malformations. This study inspired a large series of multicenter studies (Table 5) that used the KANET to assess the usefulness of this promising new scoring system for the assessment of neurological status in fetuses and the recognition of signs of early brain impairment in utero.^106,107

Table 3: Proposal for the new KANET assessment tool consisting of eight parameters

Sign	Score			Sign score
	0	1	2
Isolated head anteflexion	Abrupt	Small range (0–3 times of movements)	Variable in full range, many alternations (>3 times of movements)
Cranial sutures and head circumference	Overlapping cranial sutures	Normal cranial sutures with measurement of FX below or above the normal limit (−2 SD) according to GA	Normal cranial sutures with normal measurement of HC according to GA
Isolated eye blinking	Not present	Not fluent (1–5 times of blinking)	Fluency (>5 times of blinking)
Facial alteration (grimace or tongue expulsion)	Not present	Not fluent (1–5 times of alteration)	Fluency (1–5 times of alteration)
or
Mouth opening (yawning or mouthing)
Isolated leg movement	Cramped	Po or repertoire or small in range (0–5 times of movement)	Variable in full range, many alternations (>5 times of movements)
Isolated hand movement	Cramped or abrupt	Po or repertoire or small in range (0–5 times of movement)	Variable in full range, many alternations (>5 times of movements)
or
Hand to face movements
Fingers movements	Unilateral or bilateral clenched fist (neurological thumb)	Cramped invariable finger movements	Smooth and complex, variable finger movements
Gestalt perception of GMs	Definitely abnormal	Borderline	Normal Total score

Stanojevic et al. “Osaka Consensus Statement” DSJUOG 2011;5:4 pp. 317–329

The first application of KANET was on growth-restricted fetuses,²⁹ where mainly facial expressions and body movements were studied, and what was noticed was a decreased behavioral activity in the IUGR fetuses compared to normal growth cases. The study that followed was the first with complete neurologic postnatal assessment for all studied fetuses and according to the criteria they used neonates were divided into three groups: normal, mildly or moderately abnormal, and abnormal. According to these groups, the formation of the first KANET scoring system was decided which was as follows: 14 to 20 (normal), 5 to 13 (mildly or moderately abnormal), and 0 to 5 (abnormal) and based on this scoring system where all the following studies designed.^106,107

The first study with a large number of high-risk pregnancies identified 32 fetuses at neurological risk: they identified 7 cases with abnormal scores and 25 with a borderline KANET score. There were also 11 cases that either died in utero or had a termination of pregnancy and all of these cases had an abnormal KANET score. The 7 remaining neonates with abnormal KANET were followed up postnatally at 10 weeks of neonatal life and 3 had confirmed pathological ATNAT score. These 3 cases included a neonate with arthrogryposis, a neonate with cerebellar vermian complete aplasia, and 1 case with a history of CP in a previous pregnancy. Out of the parameters that KANET uses, facial expressions appeared to be most pathological—the fetal faces were characterized as “masks” by the authors, due to lack of expressions on 4D ultrasound. The remaining 4 pathological KANET cases had a normal postnatal assessment. These 4 cases, however, had complications of pregnancy: 1 case with ventriculomegaly, 1 case with preeclampsia, 1 case with maternal thrombophilia, and 1 case with oligohydramnios. From the 25 cases diagnosed with borderline KANET result, 22 neonates showed a borderline ATNAT score and were followed up, while the 3 remaining cases showed normal ATNAT results. An interesting paper was the one that studied a case of a fetus with prenatally diagnosed acrania. The authors studied fetal behavior and managed to document how it altered from 20 weeks of gestation onward. It was noticed that as pregnancy progressed and the control center of motoric activity shifted from the lower to the upper part, KANET score was decreasing, respectively, suggesting that neurological damage in later pregnancy is possible.⁹⁵

A study with 226 cases, including different study populations, identified 3 cases with pathological KANET scores.⁹⁸ All 3 cases had chromosomal abnormalities and all 3 of them postnatally also had an abnormal ATNAT score. Scores from antenatal KANET and postnatal ATNAT were compared between low- and high-risk groups, and showed differences between them, for 8 out of the 10 parameters—these included: head anteflexion, eye blinking, facial expressions—grimacing, tongue expulsion, mouth movements such as yawning, jawing, swallowing—isolated hand movements, hand-to-face movements, fist-and-finger movements, and GMs (Table 6).

The comparison of the two tests revealed a correlation between them, proving that the neonatal exam (ATNAT) was a satisfactory confirmation of the prenatal ultrasound examination (KANET), stating that KANET could offer useful information about the neurological status of the fetus and can be applied in clinical practice.

One of the largest studies regarding KANET,⁹⁷ including 620 cases, of both low- and high-risk populations (100 low-risk and 520 high-risk cases) showed differences in the scores between the two groups. What was interesting in this study was that most abnormal cases were noted from pregnancies with a previous history of CP (23.8%) and that most borderline scores were noted in cases with possible chorioamnionitis (56.4%). The parameters of KANET that were more notably different between the two groups were: overlapping cranial sutures, head circumference, isolated eye blinking, facial expressions, mouth movements, isolated hand movements, isolated leg movements, hand-to-face movements, finger movements, and GMs. This study confirmed the relationship of pathological KANET with increased risk of perinatal mortality and neurological impairment and showed that the results can be confirmed and are reproducible postnatally.

A recent study with a complete follow-up¹⁰⁸ postnatally up to 3 months of life, with complete postnatal documentation in all cases and showed that a normal KANET score is very reassuring of a good neonatal outcome, confirming the consistency of prenatal and postnatal assessment. Understanding the evolution of fetal movements by 4D ultrasound throughout pregnancy, and how these movements reflect the development and integrity of the fetal nervous system was a great challenge. What was shown was that during the first weeks of pregnancy the development of the frequency and the complexity of fetal movements is more important, while during the second trimester the variation of fetal movements develop, with more detailed movements (facial expressions and eye blinking) appearing at the end of this trimester. Finally, at the end of the third trimester, the number of fetal movements declines as a result of the increase of fetal rest periods, due to fetal cerebral maturation, and this is something that most pregnant women notice near term.^86,106,109 A very interesting study that tried to shed some light on the clinical dilemmas caused by the prenatal diagnosis of ventriculomegaly, compared fetuses with ventriculomegaly⁹⁹ with apparently low-risk fetuses (normal CNS appearance on ultrasound examination). A significant difference was noted between the two groups, with the KANET score decreasing as the degree of ventriculomegaly was increasing. For isolated cases of mild or moderate ventriculomegaly, no pathological KANET scores were noted and postnatal evaluation confirmed the prenatal KANET, offering valuable information for the more complete assessment of these fetuses and better counseling regarding their prognosis.

Abo-Yaqoub et al.⁹⁶ aimed to study how practical is to apply 4D ultrasonography for the assessment of fetal neurobehavior and also how useful it is for the prediction of neurological impairment. Their results showed an agreement of prenatal scores with the postnatal assessment. The parameters that were significantly different between the two groups were: isolated head anteflexion, isolated eye blinking, facial expressions, mouth movements, isolated hand movements hand-to-face movements, finger movements, and GMs. Regarding isolated leg movements and cranial sutures, the difference was not statistically significant.

Vladareanu et al.¹¹⁰ noted that the majority of normal KANET scores derived from low-risk populations that they studied, while the majority of cases with borderline or pathological KANET scores derived from the high-risk groups and in some cases were related to abnormal values of Doppler studies in IUGR fetuses. The authors concluded that KANET can be useful for the detection of neurological impairment which could become obvious during the antenatal or postnatal period.

The average KANET score was introduced for fetuses who had more than one assessment to have a more complete picture of the behavior of these fetuses. The average KANET score was derived from the mean calculation of KANET scores for each fetus throughout pregnancy since these fetuses had more than one KANET assessment. What was new from this study was the association of KANET score with fetal diurnal rhythm. For the high-risk group, 89% of the borderline scores were recorded at times that the mothers characterized them as active periods, compared with 33.3%, respectively, in the low-risk pregnancies.^111–119

Other studies^120,121 confirmed the feasibility of neurodevelopment assessment by 4D ultrasound and showed further evidence that the KANET test is useful in the early identification of fetuses prone to neurological impairment.

What was also important was to compare all parameters of KANET between high and low-risk pregnancies (Figs 7 to 10) and observe differences in fetal behavior between them. For pathological KANET score, 5 out of 8 parameters were significantly different: isolated head anteflexion, cranial sutures, and head circumference, isolated hand movement or hand-to-face movements, isolated leg movement, and finger movements.¹²² Further results showed that only high-risk patients had abnormal scores (8,5%) while comparing high- and low-risk groups it was noticed that 80.6% of high-risk patients had borderline results while 85.3% of low-risk patients were normal, both being statistically significant. For abnormal KANET results (score between 0 and 5), some were related to pregnancy complications (preeclampsia, threatened preterm labor, and drug abuse) and some were related to the fatal condition (trisomy 13, 18, and 21 and intrauterine growth restriction) previously as indicated.

While comparing Caucasian to Asian populations to check for ethnic differences, the total KANET score was normal in both populations, but there was a difference noted in total KANET scores between these two populations. When individual KANET parameters were compared, significant differences were observed in four fetal movements (isolated head anteflexion, isolated eye blinking, facial alteration or mouth opening, and isolated leg movement). No significant differences were noted in the four other parameters (cranial suture and head circumference, isolated hand movement or hand to face movements, fingers movements, and gestalt of GMs), showing that ethnicity is a parameter that should be considered while evaluating fetal behavior, especially during the assessment of fetal facial expressions. The authors concluded that although there was a difference in the total KANET score between Asian and Caucasian populations, all the scores in both groups were within the normal range proving that ethnical differences in fetal behavior do not affect the total KANET score, but close follow-up should be continued in some borderline cases.¹²³

Unpublished data from Greece, from 655 singleton pregnancies, showed that KANET is a method that is feasible in everyday clinical practice, with a success rate of 95% and a very low negative predictive value. For the cases that KANET could not be completed, the reason was severe oligohydramnios, fibroid uterus (difficult imaging), very increased BMI, and a case that due to vasovagal reaction-supine hypotensive syndrome ultrasound examination could not be completed. From the 655 cases, 1,712 KANET were performed from only two operators and the interobserver variability was calculated showing adequate results for all parameters, with the lowest being for facial alterations (k-value = 0.68) and the highest for finger movements (k-value = 0.84). This study was primarily designed to compare the neurological status of pregnancies complicated by diabetes, compared to low-risk pregnancies and it did show that there was a difference between the fetal neurobehavior of these two groups, with the diabetic pregnancies having lower scores.¹²⁴

Figures 11 to 13 are illustrating important parameters of KANET depicted by high definition (HD) 4D US.

Interpretation of KANET Test esearch

One of the greatest challenges in perinatal medicine is the assessment of fetal neurobehavior and detection of fetal neurological impairment in utero. KANET is the first method that applied 4D ultrasound for the assessment of the fetus in the same way that a neonate is assessed neurologically after birth by neonatologists and it appears to be a strong diagnostic method for the detection of neurological impairment and for the assessment of fetal neurobehavior, conditions that were inaccessible with the traditional prenatal diagnostic methods used so far.⁷⁴ Studies have proved the validity of this method,^27,28,86 its applicability in everyday clinical practice, especially for high-risk cases, how and by whom it should be performed, and what is the value of the result of KANET and how it should be managed. Diagnosis of neurological impairment prenatally is very difficult and usually, all these diagnoses are made postnatally, even months or years after delivery. What is more neurological conditions, such as CP are not adequately understood and falsely attributed to incidents during labor, although it has been proven that the majority of CP cases originate sometime during in utero life and are not related to intrapartum events. All these things lead to delayed diagnosis of neurological conditions. The later a neurological impairment is diagnosed the less is the possibility of an effective intervention. It would be extremely challenging to have a timely diagnosis of such conditions, even during in utero life, to increase the possibility of an effective intervention or even treatment. KANET offers the possibility of prenatal detection of fetuses at risk for neurological problems, offering the possibility of even an in utero intervention or at least an early postpartum intervention.¹²⁵ The earliest physiotherapy is commenced and intervention programs are applied in neonates that are born prematurely or with neurological problems the better the neurodevelopmental outcome of these neonates, with the cognitive benefits persisting into pre-school age. KANET appears to be able to offer this advantage of early identification of these fetuses with neurological problems, so that they could be put under treatment as early as possible, aiming for a better outcome.^120,126

What is more, the explicitly detailed pictures obtained by the new ultrasound machines but also the advanced techniques of molecular genetics, many times brings us, as ultrasound specialists, across findings (anatomical and chromosomal) of uncertain clinical significance and prognosis, especially regarding the neurological integrity of the fetus.^127–129 A method like KANET offers a more comprehensive diagnostic approach to such dilemmas and hopefully soon with more data we could form a complete neurosonobehavioral assessment of the fetus and more complete counseling of these couples.¹³⁰

KANET has currently been introduced in everyday clinical practice by many centers for the assessment of fetal neurobehavior of not only high-risk cases but also low-risk pregnancies. Studies show that the sensitivity and specificity of the test are satisfactory, as are the positive and negative predictive values and the inter- and intraobserver variability of this method. The KANET has been introduced into systematical training and already ultrasound specialists have been certified to perform this examination. Hopefully, the application of KANET on larger populations, both high- and low-risk, will give more knowledge regarding early detection of fetuses at risk for neurological impairment, allow accurate diagnosis prenatally, and as a consequence prompt intervention that could possibly improve the outcome of some of these neonates.