Table e Demographics for the neonates with and without wmi



Download 2.29 Mb.
Date17.11.2017
Size2.29 Mb.
Table e-1. Demographics for the neonates with and without WMI




WMI (n=58)

No WMI (n=124)

p

Maternal level of education Median [IQR]

5 [5-6]

5 [5-6]

0.2

Males, No. (%)

26 (44.8%)

69 (55.6%)

0.2

Birth gestational age, weeks Median [IQR]

28.6 [26.3-29.8]

27.9 [26-30]

0.8

PMA at MRI, weeks Median [IQR]

32 [30.6-33]

32.4 [30.6-34.1]

0.4

Head circumference at birth, cm Median [IQR]

26.3 [24.4-27.6]

26 [23.5-27.5]

0.2

SNAP-PE Median [IQR]

9 [0-27]

17 [0-37]

0.3

Hypotension (%)

22 (37.9%)

43 (34.7%)

0.7

Culture positive infection (%)

17 (29.3%)

35 (28.2%)

0.9

Patent ductus arteriosus (%)

28 (48.3%)

53 (42.7%)

0.8

Chronic lung disease (%)

9 (15.5%)

24 (19.0%)

0.5

Necrotizing enterocolitis (%)

8 (13.8%)

27 (21.8%)

0.2

Intraventricular haemorrhage (IVH) (%)







0.9

Grade 1-2

31 (53.4%)

41 (33.1%)

Grade 3-4

2 (3.4%)

3 (2.4%)

Ventriculomegaly (%)







0.8

Grade 1

15 (25.9%)

18 (14.5%)

Grade 2

3 (5.2%)

3 (2.4%)

Cerebellar haemorrhage (CH) (%)










0

51 (87.9%)

109 (87.9%)+

0.99

1

7 (12.1%)

14 (12.1%)


+Missing a data point for one subject with CH-0

Table e-2. Neurodevelopmental outcome assessed with the Bayley-III




No WMI (n=124) Median [IQR]

WMI-1 (n=30) Median [IQR]

WMI-2 (n=17) Median [IQR]

WMI-3 (n=11) Median [IQR]

P*

Age at follow up, months

18.6

18.6

18.6

19

0.24

[18.3-19.2]

[18.3-19.2]

[18.4-20.8]

[18.3-19.8]

Bayley-III motor

100

99

100

82

0.001

[91-107]

[88-108]

[88-103]

[75-87]

Bayley-III cognitive

105

108

110

100

0.26

[100-115]

[101-110]

[95-110]

[88-110]

Bayley-III language

100

102

106

94

0.66

[86-109]

[89-111]

[94-112]

[83-111]

WMI, white matter injury, Bayley-III, Bayley Scales of Infant Development, 3rd edition, IQR, interquartile range.

* ANOVA

Includes two neonates with macro-cystic lesions

Figure e-1. Manually segmented WMI on an early-in-life T1-weighted image.



Figure e-1. WMI segmentation for each neonate was performed with simultaneous coronal, axial, and sagittal views of the brain using Display software (http://www.bic.mni.mcgill.ca/ServicesSoftwareVisualization). The purple cross is at the same location in the neonatal brain on the three views.

Figure e-2. Segmented brain lobes of the neonatal brain template.

Figure e-2. Segmented lobe atlas in coronal (left), axial (inferior), axial (superior) and sagittal (right) views. The four lobes (frontal, parietal, temporal, occipital) of the entire cerebral cortex and the basal ganglia/thalamus were manually delineated separately in each hemisphere by one of the investigators (EGD) using the software program Display, part of the minc tools software package (http://www.bic.mni.mcgill.ca/ServicesSoftwareVisualization). The frontal lobes included all grey and white matter from the central sulcus posteriorly to the anterior most point of the frontal pole and inferiorly to the lateral sulcus. Included in the frontal lobes were the entire medial, lateral and orbital surfaces. The rostrum, genu and body of the corpus callosum were also included in the frontal lobes masks as was the grey matter of the insular cortices. The parietal lobes included the entire grey and white matter territory posterior to the central sulcus extending inferiorly to include the angular gyrus. The posterior most point of the parietal lobe was defined by the parieto-occipital sulcus. The parietal lobe included the superior and inferior parietal lobules, the precuneus, posterior cingulate (posterior to the central sulcus) and splenium of the corpus callosum. The temporal lobes were comprised of the grey and white matter inferior to the lateral sulcus. The anterior border was the temporal pole and the inferior border was the territory of the posterior ramus of the lateral sulcus extending down to the region of the preoccipital notch of the inferior temporal sulcus. The temporal lobes included the medial temporal lobe territory on the ventral surface of the brain. The occipital lobes including all grey and white matter posterior to the parieto-occipital sulcus extending down to the preoccipital notch territory to include the cuneus and the lateral surface of occipital cortex. Lastly, the entire basal ganglia (striatum, globus pallidus) and thalamus were segmented to include all the grey matter of the nuclei in the associated structures. Additionally, just lateral to the putamen, the white matter extending to the territory of the extreme capsule was also included in the basal ganglia/thalamus masks.



Supplemental Material

  1. Intra- and inter-rater reliability of manual segmentation: The intra-rater reliability of WMI manual labelling was evaluated on 15 images for each rater and inter-rater reliability was estimated on 5 images labelled by both raters. To ensure that rater memory did not play any confounding role in the re-labelling, the initial and re-segmentations were performed with intervals of longer than 1 month. A commonly used method for similarity assessment, the Dice similarity coefficient (DSC) or Dice’s Kappa, was employed to assess the agreement between the original and the re-segmented WMI volumes for the protocol reliability test:

where is the common voxels of the original and re-segmented WMI volumes, and is the sum of the voxels that are unique to each of the segmentations. A perfect agreement between the two segmentations results in (Guo et al., 2015).




  1. Definition of brain lobes in the early-preterm brain template: The frontal lobes included all grey and white matter from the central sulcus posteriorly to the anterior most point of the frontal pole and inferiorly to the lateral sulcus. Included in the frontal lobes were the entire medial, lateral and orbital surfaces of the hemispheres. The rostrum, genu and body of the corpus callosum were also included in the frontal lobes. The parietal lobes included the entire grey and white matter territory posterior to the central sulcus extending inferiorly to include the angular gyrus. The posterior most point of the parietal lobe was defined by the parieto-occipital sulcus. The parietal lobe included the superior and inferior parietal lobules, the precuneus, posterior cingulate (posterior to the central sulcus) and splenium of the corpus callosum. The temporal lobes were comprised of the grey and white matter inferior to the lateral sulcus. The anterior border was the temporal pole and the inferior border was the territory of the posterior ramus of the lateral sulcus extending down to the region of the preoccipital notch of the inferior temporal sulcus. The temporal lobes included the medial temporal lobe territory on the ventral surface of the brain. The occipital lobes including all grey and white matter posterior to the parieto-occipital sulcus extending down to the preoccipital notch territory to include the cuneus and the lateral surface of occipital cortex.




  1. Development of odds ratio maps: The odds ratio of having adverse developmental outcomes at each voxel can be derived using the following formula:

Where is the number of neonates with adverse outcome who have WMI at a specific voxel, is the number of neonates with typical outcome who have WMI at a specific voxel, is the number of neonates with adverse outcome who do not have WMI at a specific voxel, and is the number of neonates with typical outcome who do not have WMI at a specific voxel. To prevent divisions by zero, we set the voxels that exhibited lesions in the adverse outcome group but with no lesions in the typical outcome group to a value of 1.









Share with your friends:


The database is protected by copyright ©dentisty.org 2019
send message

    Main page