Primary Objective: to develop and evaluate a machine learning (ML) prediction model of brain injury in neonates at high-risk for CP to predict motor, behavioral, and cognitive outcome more accurately.This aim will be achieved by: developing and…
ID
Source
Brief title
Condition
- Congenital and peripartum neurological conditions
Synonym
Research involving
Sponsors and support
Intervention
Outcome measures
Primary outcome
The creation of a ML algorithm (including standard clinical care assessment:
MRI, EEG, and clinical data collected up to 3 months of age, including GMA and
HINE) being able to early predict motor, behavioral, cognitive outcome. The
creation and implementation of automated tools to analyze these early
assessments in at-risk infants to obtain a more precise, individualized
diagnosis and prognosis of CP based on harmonized scans and protocols (GMFCS,
BSID-III, CBCL).
Secondary outcome
The development of recommendations to support parents in the process of
disclosure of diagnosis and communication around prognosis of the development
of their child, using questionnaires and interviews with parents.
Background summary
Cerebral palsy (CP) is the most common cause of physical disability in children
but is still diagnosed too late. Consequently, many children with CP do not get
specific intervention until their second birthday, which has consequences for
motor and cognitive outcomes, as the great part of the entire neuroplastic
window for motor learning is misspent. Specific and reliable tools for the
early detection of infants with CP have been recently defined and are now part
of the first International Clinical Practice Guidelines. Infants with perinatal
risk factors for CP can reliably receive an early diagnosis of CP before 6
months using a combination of brain magnetic resonance imaging (MRI) and either
the general movement assessment (GMA) and/or the Hammersmith infant
neurological examination (HINE). In addition, electroencephalography (EEG) is
essential to define and monitor the level of brain maturation and seizures in
high-risk newborns, thus contributing to outcome prediction. As CP is a
heterogeneous condition, it is critical to be able to formulate and communicate
1) an early diagnosis and 2) the functional prognosis of the child at a very
young age, so clinicians and families are informed and can make informed
decisions on treatment goals and (novel) interventions. Nowadays the
combinations of early assessments in one prediction model is lacking, as well
as the precision in prediction.
Study objective
Primary Objective:
to develop and evaluate a machine learning (ML) prediction model of brain
injury in neonates at high-risk for CP to predict motor, behavioral, and
cognitive outcome more accurately.
This aim will be achieved by: developing and validating a ML prediction model
that will include the following well-established diagnostic tools: MRI, EEG,
GMA and HINE combined with the clinical variables during admission in infants
with MRI diagnosed brain injury with high risk of CP. Next to automated scoring
tools for each modality.
Secondary Objective(s):
1. a. To gain insight in parental mental well-being and family- and social
functioning over the first two years after the birth of their child with brain
injury, born preterm or full-term, at high risk of developing CP.
1. b. To examine the relation of parents* experiences around the disclosure of
diagnosis, child-related factors and social support with parental mental
well-being.
2. To understand the needs and preferences of parents related to the process of
disclosure of diagnosis and related to the information they receive regarding
prediction and prognosis, to identify what information is meaningful to
families.
Study design
This is a longitudinal, prospective, observational multicenter study where
several populations of newborn infants with brain injury and high-risk of CP
will be enrolled.
We will collect multimodal data from a unique cohort of ~1000 newborn infants
with brain injury who are at high risk of developing CP; we will focus on
preterm and full-term infants with brain injury at high-risk for CP confirmed
using neonatal MRI, please refer to the "inclusion criteria" in the protocol
for elaboration.
These infants will be enrolled from 8 large EU neonatal centers with neonatal
neurology expertise and followed closely from birth, including initial
neuroimaging (MRI) and neurophysiologic assessment using aEEG and EEG, followed
by behavioral, clinical, and neuropsychological evaluations through infancy up
to 2 years of age (within this 5-year project). Parents also have the option
to have their child's crying sounds automatically recorded, and to opt for a
50-minute autism spectrum disorder test for their child at 2 years old. Please
refer to "Figure 1" in the added protocol.
The project will start in January 2023: 2,5 years of inclusion, followed by 2
years of follow-up of the infants .
After this 5-year project, we will continue to follow these children in order
to define the prognosis regarding cognitive and behavioral domains; this
long-term follow-up is particularly important, given that motor deficits are
not the main factor that limits the child*s social participation, schooling,
and employability.
Study burden and risks
Most of the data for this study are obtained from standard clinical care
procedures. For the child, all data are collected from procedures done as part
of standard clinical care. For the parents, however, there is an additional
action outside of clinical care, namely completing two sets of questionnaires
when their child is 4 months and 24 months old. These sets of questionnaire
take the parent 20 minutes and 30 minutes respectively to complete. The burden
and risks of participating in this study are very low. On the other hand, early
prediction of CP by the ML model could improve long-term motor, cognitive and
behavioural outcomes in future children. The questionnaires may improve future
parental support.
Lundlaan 6
Utrecht 3584 EA
NL
Lundlaan 6
Utrecht 3584 EA
NL
Listed location countries
Age
Inclusion criteria
-All infants with confirmed brain injury on MRI at high risk for cerebral palsy.
-Written informed parental consent (Dutch, English, French, German, Italian,
Spanish).
Exclusion criteria
-Infants not matching the inclusion criteria
-Any proven or suspected severe congenital anomaly, genetic or metabolic
disorder
-Presence of an infection of the central nervous system
-Parents < 18 years old
-Not being able to read one of the six Informed Consent languages
Design
Recruitment
Followed up by the following (possibly more current) registration
No registrations found.
Other (possibly less up-to-date) registrations in this register
No registrations found.
In other registers
Register | ID |
---|---|
CCMO | NL83183.041.22 |