Student: Daniel Grzech
1st supervisor: Bernhard Kainz, Imperial College London
2nd supervisor: Jo Hajnal, King’s College London
Neonatal movements are commonly assessed as a way of predicting neurological function of the baby, but methods currently in use are subjective, time-consuming and are not standardised. It has been shown that there is continuity between fetal and neonatal movement patterns, and that fetal movements are affected by certain neurological disorders, but fetal movement patterns are not routinely assessed. This project will develop a system for automatically tracking and characterising fetal movements visualised using cine MRI. A large bank of fetal movement data has already been gathered for normal subjects and for subjects at increased risk of neurological conditions, e.g., ventriculomegaly (enlarged lateral ventricle in the brain). From this data, this project will identify one or more movement-based ‘biomarkers’ indicative of neurological function using image-based feature identification and tracking methods combined with machine learning approaches. This project forms part of a platform approach to objective assessment of fetal and neonatal movement patterns for early diagnosis of neurological abnormalities.
Assessment of neonatal movements is commonly performed in infants at higher risk of neurological abnormalities (e.g., preterm, idiopathic respiratory distress syndrome, early CNS insults) as a predictive measure of neuromotor function . Correlations have been reported for movement parameters and outcomes such as cerebral palsy and attention-deficit hyperactivity disorder [2, 3]. However, there are several major problems with the current methodologies for assessing neonatal movements. Firstly, they are subjective and have been shown to be assessor-dependant . Secondly, they are time-consuming, involving extended video recording of the infant during an active phase and further time spent on analysing the footage. Thirdly, different hospitals and clinicians use different methods to quantify and characterise the movements . There is evidence that fetal movements are abnormal in cases of certain neurological disorders, such as anencephaly, trisomy 21 and subependymal heterotopia. However, fetal movements are not routinely assessed (despite evidence of continuity between fetal and neonatal movement patterns ), and no objective means of characterising fetal movement exists, meaning that a potential tool for prenatal diagnosis of the presence or severity of a neurological condition is not being utilised. Our team is developing a platform approach to automated assessment of fetal and neonatal movements for identification of neurological health and function across multiple imaging modalities including MRI and ultrasound of fetuses and video recordings of neonates. This project is the first step towards such a platform approach, and will use already acquired banks of cine MRI fetal movement data of normal controls and subjects at increased risk of neurological abnormalities.
The recent emergence of high-resolution fetal magnetic resonance cine imaging has made it possible to obtain a large field of view of the active fetus at a range of gestational ages. We will use an existing large (n=100) bank of cine MRI data at KCL which show snapshots of fetal movements for a range of gestational ages (20-38 weeks) with sub-populations of subjects at increased risk of neurological abnormalities and normal controls. Ethical approval and intellectual properly agreements are already in place to share these data with the project team.
The proposed project will perform a rigorous analysis of fetal movements in normal and at-risk sub-populations in order to propose and validate biomarkers for fetal neurological function that will serve as a diagnostic and prognostic clinical tool. In pilot work, we have demonstrated feasibility of the methodology, having a) successfully tracked fetal movements from cine MRI data , and b) shown that the speed and estimated energy of fetal kicks is different between normal fetuses and fetuses diagnosed with a neurological abnormality [unpublished data]. This project will build upon our pilot work, focussing on retrospective cine MRI data from fetuses with trisomy21, subependymal heterotopia, and ventriculomegaly, which has been identified as a risk factor for neurodevelopmental disorders including autism .
The impact of this project will be that it will be the first to perform automated analysis of fetal movements for a large dataset, and the first to propose and test a new means of prenatally diagnosing fetal abnormalities and assessing the severity of neurological complications at an early stage of pregnancy. This project is achievable, as feasibility has been demonstrated through our pilot studies, and we have guaranteed access to the necessary data.
1. Prechtl, Heinz FR. “General movement assessment as a method of developmental neurology: new paradigms and their consequences The 1999 Ronnie MacKeith Lecture.” Developmental Medicine & Child Neurology 43.12 (2001): 836-842.
2. Hadders-Algra, Mijna, and Annelies MC Groothuis. “Quality of general movements in infancy is related to neurological dysfunction, ADHD, and aggressive behaviour.” Developmental Medicine & Child Neurology 41.6 (1999): 381-391.
3. Adde, Lars, et al. “General movement assessment: predicting cerebral palsy in clinical practise.” Early human development 83.1 (2007): 13-18.
4. Harris, Susan R., et al. “Reliability of observational measures of the Movement Assessment of Infants.” Physical therapy 64.4 (1984): 471-475.
5. Spittle, Alicia J., Lex W. Doyle, and Roslyn N. Boyd. “A systematic review of the clinimetric properties of neuromotor assessments for preterm infants during the first year of life.” Developmental Medicine & Child Neurology 50.4 (2008): 254-266.
6. Almli, C. Robert, Robert H. Ball, and Mark E. Wheeler. “Human fetal and neonatal movement patterns: Gender differences and fetal-to-neonatal continuity.” Developmental Psychobiology 38.4 (2001): 252-273.
7. Verbruggen, Stefaan W., et al. “Modeling the biomechanics of fetal movements.” Biomechanics and modeling in mechanobiology (2015): 1-10.
8. Gilmore, John H., et al. “Outcome in children with fetal mild ventriculomegaly: a case series.” Schizophrenia research 48.2 (2001): 219-226.