A Neonatal Bimodal MR-CT Head Template

Completed

• Conditions: Magnetic Resonance Imaging
• Interventions: Other: CTscan and MRI images

• Registration Number: NCT03674034
• Lead Sponsor: Centre Hospitalier Universitaire, Amiens

Brief Summary

This is a retrospective study based on the images contained in the database of the University Hospital of Amiens. There is no further examination. Electroencephalography (EEG) and Near-Infrared Spectroscopy (NIRS). These sources are widely used to specify the source of certain evoked potentials or to better define the sources of epileptic graphoelements, especially in the pre-surgical assessment of refractory epilepsies.

It is therefore important that these regions are more widely available in the constitution of the world, and that they are suitable for use in the design of new technologies. Although MRI is the gold standard for soft tissue segmentation, it is not suitable for bone extraction; especially in the newborn or the bone is very thin. On the other hand, CTscan is an excellent method for extracting bone. The contrast between bone and soft tissue is excellent. In this case, with the CTscan, the fontanelles are identified as discontinuities between the images of the temporal scales.

Detailed Description

Magnetic resonance imaging (MRI) and brain scan (CT Scan) are the two main modalities of structural brain imaging that are widely used in routine practice. The extraction of bone from MRI and CTscan images is an important step in the segmentation of the different tissues of the subjects' heads. It is indeed necessary to know precisely the electrical characteristics (conductivities) and optical (absorption, diffraction) of tissues (skin, bone, LCS, gray matter and white matter), or by electric currents or by photons. One of the major problems is the one that is not very poorly visualized in MRIs in CTscan, the bone is well individualized and that inversely the various brain structures are difficult to extract. This justifies the fact of performing a coregistration of images obtained in MRI and CTscan.

The skull of the newborn is an inhomogeneous structure. The fontanelles are invisible in MRI and identifiable in CTscan.

Electroencephalography (EEG) and Near-Infrared Spectroscopy (NIRS). These sources are widely used to specify the source of certain evoked potentials or to better define the sources of epileptic graphoelements, especially in the pre-surgical assessment of refractory epilepsies.

It is therefore important that these regions are more widely available in the constitution of the world, and that they are suitable for use in the design of new technologies.

Although MRI is the gold standard for soft tissue segmentation, it is not suitable for bone extraction; especially in the newborn or the bone is very thin. On the other hand, CTscan is an excellent method for extracting bone. The contrast between bone and soft tissue is excellent. In this case, with the CTscan, the fontanelles are identified as discontinuities between the images of the temporal scales.

Study Timeline

• Study Start: 2016-01-01
• Primary Completion: 2016-01-01
• Study Completion: 2017-01-01
• Status Verified: 2018-09
• Study First Submitted: 2018-09-14
• Study First Submitted QC: 2018-09-14
• Last Update Submitted: 2018-09-14
• Study First Posted: 2018-09-17
• Last Update Posted: 2018-09-17

Recruitment & Eligibility

• Status: COMPLETED
• Sex: Male
• Target Recruitment: 150

Inclusion Criteria

• children aged between 0 and 2 months new born at term
• with no structural abnormalities

Exclusion Criteria

• infants > 2 months

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Arm && Interventions

Group	Intervention	Description
two months	CTscan and MRI images	50 CTscan and MRI images of children aged two months
one month	CTscan and MRI images	50 CTscan and MRI images of children aged one month
term	CTscan and MRI images	50 CTscan and MRI images of children aged at term

Primary Outcome Measures

Name	Time	Method
CTscan images	2 months	The objective of this study is to evaluate a model-based approach for extracting fontanel from CTscan images. Then, recalibrate these CT scan images in a unique repository with MRI images to obtain a composite image of different brain tissues such as brain, LCS and skin (MRI) and bone and fontanelle (CTscan).