Local shape analysis of the thalamus in extremely preterm born young adults
Eliza Orasanu1, Andrew Melbourne1, Zach Eaton-Rosen1, David Atkinson2, Joshua Lawan3, Joanne Beckmann4, Neil Marlow4, and Sebastien Ourselin1

1Translational Imaging Group, Centre for Medical Image Computing, University College London, London, United Kingdom, 2University College London, London, United Kingdom, 3University College Hospital, London, United Kingdom, 4Institute for Women's Health, University College London, London, United Kingdom

Synopsis

Alterations of thalamic structures may cause disruptions in thalamic-cortical-thalamic circuitry and affect cognition. In this work we present a local shape analysis of the thalamus in extremely preterm born young adults when compared to their term born peers. We perform a groupwise shape analysis after spectral matching registration. After correcting for gender and thalamic volume, it resulted that the anterior and superior thalamic regions, connected to regions responsible for executive function, working memory, language and verbal memory, show most shape variations.

Introduction

The last trimester of pregnancy is a period of major brain development, with changes in volume, appearance and connectivity of the foetal brain. Birth before 27 weeks (extremely preterm) implies that this development will take place outside the mother’s womb and these infants are prone to increased rates of adverse neurological outcome [1]. The thalamus is an important part of the brain and alterations of the thalamic structure are likely to cause disruption in the thalamic-cortical-thalamic circuitry and thus may affect cognitive performance. It has been shown that the size and structure of the thalamus are affected by preterm birth [2] and we hypothesis that these difference may persist into adulthood. Mapping differences between extremely preterm and term born adults can provide us with an understanding of the long-term structural impact of extreme prematurity.

Methods

T1-weighted MR data was acquired from 62 young adults born at <26 weeks of gestation (38 females + 24 males) and 47 control subjects (28 females + 19 males). All subjects were 19 year old adults born in 1995. We segmented the thalamus of each subjects using the GIF framework [3]. Thalamus segmentations were used to obtain smooth triangle-based meshes of the thalamus. For each group (preterm and control) we chose a random subject as template, to which we mapped all other surfaces using joint spectral matching with a CPD initialisation. The mappings were used to create a mean shape for each of the two groups. Morphological changes between the groups were then investigated by computing the difference in vertex position (displacement maps) of the mean shapes thalamic surfaces after another step of joint spectral matching. This pipeline was repeated to investigate differences between the groups by gender. Groupwise shape analysis using spectral matching has previously been shown to give reliable results [4]. We used the Hotelling T2 two sample metric to derive a local group difference metric and local statistical p-values for all corresponding points [5]. After fitting a multivariate general linear model to our data, correcting for thalamic volume and gender, we computed the vertex-wise T-statistics using a random field theory multiple-comparison correction to yield an equivalent p-value of 0.05 and we then generated a map of group difference.

Results

Firstly, we notice that the mean thalamic volume is larger in control subjects (12.46±1.17 cm3) than in the preterm population (10.46±1.22 cm3). Secondly, the local shape differences between control and preterm groups are larger in the anterior part of the thalamus than in the posterior part (Figure 1). Furthermore, these differences are larger in the left hemisphere than in the right.

When separating the subjects by gender, we notice that the mean thalamic volume is smaller in females than males: 10.29±1.17 cm3 and 10.80±1.21 cm3 in the preterm population for female and male, respectively and 11.29±1.03 cm3 and 13.28±0.92 cm3 for the control population for female and male, respectively. Local shape differences between the preterm and term females are quite constant with no noticeable asymmetry (Figure 2). The local shape differences between the preterm and term males are large, especially in the anterior part of the thalamus (Figure 3). The differences are greater in the left hemisphere than in the right.

The group shape significance map, corrected for thalamic volume and gender (Figure 4), shows that the shape of the superior-lateral part of the thalamus is significantly different between the preterm and control groups, with a slight non-significant left-right asymmetry.

Discussion/Conclusion

We investigated differences in thalamus shape and volume between preterm and term-born young adults by performing a group comparison on vertex displacement, with matching carried out using joint spectral matching. The thalamus was smaller in preterm-born individuals with differences mainly in the anterior thalamus and these differences were more pronounced in male subjects. The anterior thalamus connects to brain regions with roles in executive function, working memory, problem solving, mood and motivation [6]. Thalamic shape differences were more pronounced in the superior part, which has substantial connections to the temporal lobe [7], with role in language and verbal memory. Similar anterior and superior thalamic differences have been found in studies on ADHD subjects [7] and it may be interesting to explore the links further between thalamic shape, composition and the influence on an extreme prematurity-ADHD correlation [8]. Our future work will explore this possibility making use of the results of a neurocognitive assessment battery.

Acknowledgements

We would also like to acknowledge the MRC (MR/J01107X/1), the National Institute for Health Research (NIHR), the EPSRC (EP/H046410/1) and the National Institute for Health Research University College London Hospitals Biomedical Research Centre (NIHR BRC UCLH/UCL High Impact Initiative- BW.mn.BRC10269). This work is supported by the EPSRC-funded UCL Centre for Doctoral Training in Medical Imaging (EP/L016478/1).

References

[1] N. Marlow, D. Wolke, M. A. Bracewell, and M. Samara, “Neurologic and Developmental Disability at Six Years of Age after Extremely Preterm Birth,” N. Engl. J. Med., vol. 352, no. 1, pp. 9–19, 2005.

[2] J. P. Boardman, C. Craven, S. Valappil, S. J. Counsell, L. E. Dyet, D. Rueckert, P. Aljabar, M. Rutherford, T. M. Chew, J. M. Allsop, F. Cowan, and D. Edwards, “A common neonatal image phenotype predicts adverse neurodevelopmental outcome in children born preterm.,” Neuroimage, vol. 52, no. 2, pp. 409–14, Aug. 2010.

[3] M. J. Cardoso, M. Modat, R. Wolz, A. Melbourne, D. Cash, and D. Rueckert, “Geodesic Information Flows?: Spatially-Variant Graphs and Their Application to Segmentation and Fusion,” IEEE Trans. Med. Imaging, vol. 34, no. 9, pp. 1976–1988, 2015.

[4] M. Shakeri, H. Lombaert, S. Lippé, and S. Kadoury, “Groupwise shape analysis of the hippocampus using spectral matching,” in SPIE Medical Imaging, 2014.

[5] M. Styner, I. Oguz, S. Xu, D. Pantazis, and G. Gerig, “Statistical Group Differences in Anatomical Shape Analysis using Hotelling T2 metric,” in Proceedings SPIE 6512, 2007, p. 65123 z1–z11.

[6] E. J. Hughes, J. Bond, P. Svrckova, A. Makropoulos, G. Ball, D. J. Sharp, A. D. Edwards, J. V Hajnal, and S. J. Counsell, “Regional changes in thalamic shape and volume with increasing age,” Neuroimage, vol. 63, no. 3, pp. 1134–1142, 2012.

[7] S. Xia, X. Li, A. E. Kimball, M. S. Kelly, I. Lesser, and C. Branch, “Thalamic shape and connectivity abnormalities in children with attention deficit / hyperactivity disorder,” Psychiatry Res. Neuroimaging, pp. 1–7, 2012.

[8] T. M. O'Shea, L. C. Downey, and K. K. C. Kuban, “Extreme prematurity and attention deficit: epidemiology and prevention,” Front. Hum. Neurosci., vol. 7, no. September, pp. 1–5, 2013.

Figures

Figure 1. Local shape differences between the preterm and term mean thalamus: superior aspect and lateral aspect (left and right). Units of scale are mm.

Figure 2. Local shape differences between the female preterm and female term mean thalamus: superior aspect and lateral aspect (left and right). Units of scale are mm.

Figure 3. Local shape differences between the male preterm and term mean thalamus: superior aspect and lateral aspect (left and right). Units of scale are mm.

Figure 4. Significance of thalamus group difference controlling for thalamic volume and gender. P-value of 0.05 corresponds to a T-stat of 4.3125, hence regions with T-stat values greater than 4.3125 will pass the random field theory based multiple comparison thresolding at 0.05 significance level.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0565