Tract-Based Spatial Statistics - Essay Example

Tract-Based Spatial Statistics Introduction Tract-based spatial statistics (TBSS) is a method used in analyzing diffusion tensor imaging data (Keihaninejad, Ryan, Malone, Modat, Cash, et al. 2012 p 459). It is important in the study of white matter within the brain during the process of its development. The target of the tract-based spatial statistics is the difference in white matter voxels that have high fractional anisotropy (FA) in the place of main fiber tracts by recording all the subjects to one reference and the development of a fractional anisotropy skeleton. Analyzing diffusion weighted MRI (DWI) data is currently very critical in neuro-imaging studies. DWI has information needed in the assessment of the integrity of white matter, its connectivity and architectural patterns (Wang, Gupta, Liu, Zhang, Escolar, et al. 2011 p1580). Tract-based spatial statistics approach works automatically in the analysis of DTI where a method based on voxel-wise comparison is used to analyze associations between various subjects for example the differences existing between different groups. TBSS may be a good method for brain analysis but its disadvantages are also evident and this has been proved by many researchers. The focus of this paper is to discuss the various studies done in this field, the areas where TBSS is used and its advantages and disadvantages (Seeley, Crawford, Zhou, Miller, Greicius 2009 p50). Studies in this field Several studies have been made in this field. Among these studies is the one done by Shiva Keihadinejad and others in the Dementia Research Centre, University College London Institute of Neurology about Tract-based spatial statistics (TBSS). In their study, they considered the effect that the choice of reference has within the TBSS pipeline which may either be a standard template or a particular subject in the study or even a group wise average or a study specific template (Keihadinejad et al., 2012 p. 450). The authors seek to show the effect that choice of reference has on sensitivity is simulation studies, specificity, and real comparison of patients with Alzheimer’s disease to controls. In the first and second options above simulated FA decreases and deformations were used on control subjects for purposes of simulating changes of WM integrity and shape. This can be likened to what is observed in patients with Alzheimer’s disease so that it can provide facts for the evaluation of the different TBSS reference techniques (Smith, Jenkinson, Johansen-Berg, Ruekert, Nichols, Mackay, Watkins, Ciccarelli, Cader, Matthews, and Behrens, 2006 p 501). Another study done in this field is the one done by Douaud et al in which he wanted to handle huge enlargement in ventricles in Alzheimer’s disease patients. In his research, he created an FA template specific to his study through the registration of all the native space FA images to another FA template inside the MNI space and later getting their average. Original scans of Fractional Anisotropy were not registered linearly to this study specific template of fractional anisotropy and therefore the study relied on the registration performance and not the atlas. Another study is the one done by Zalesky in which he wanted to make valid the skeleton projection algorithm in the core of TBSS. He developed a methodology for evaluation whereby he used synthetic warps of an image of ground truth to evaluate in a quantitative manner, the performance of TBSS on the basis of 3 healthy people and two different FA image sets. Form his study, Zalesky found that even if the skeleton projection may recover only a tenth of its misalignment after registration it still ended in a smaller error of the value of fractional anisotropy compared to the use of Gaussian smoothing to minimize mis-registration effects. Another study was done by Kieseppä et al., on “Major depressive disorder and white matter abnormalities: a diffusion tensor imaging study with tract-based spatial statistics” (Douaud, Jbabdi, Behrens, Menke, Gass, et al. 2011 p 885). In their study they examined the white matter structure in the whole brain with DTI in patients with major depressive disorders and in their middle ages. They used novel tract-based spatial statistics. The authors discovered that in comparison to controls, the patients suffering from major depressive disorders displayed a trend with lower fractional anisotropy values in the left side sagittal stratum and lower fractional anisotropy within the right side cingulated and the front part of corpus callosum. Their conclusion was that regressing out the period taken by severity of disorder within the model had no effect on the sagittal stratum results but had a dissipating effect on the FA in latter regions (Smith et al., 2006 p 1489). In another study, Ball G. et al studied the optimized tract-based spatial statistics protocol for neonates. They applied it to chronic lung disease and prematurity. In this study, the authors studied the effects of chronic lung disease on white matter anisotropy and diffusivity by use of optimized protocol. Chronic lung disease is happens to be a risk factor that is independent and it relates to the abnormal development of white matter (Seeley et al., 46). Where TBSS is used Tract-based spatial statistics (TBSS) is a tool that has proved to be very critical in studying white matter in the process of development by the use of DTI, although standard TBSS protocols present limitations in neonatal studies (Douaud et al., 2011p 887). White matter is a section of the human brain and any malfunctioning in its development results in many complications in the brain and body. In the recent past, there has been high interest in the use of magnetic resonance diffusion imaging in the provision of information about the brain and especially on anatomical connectivity. This is done through the determination of the anisotropic diffusion of the water in tracts inside the white matter. Fractional anisotropy is among the measures that are often obtained from diffusion data which helps to put in quantitative terms the strength in the direction of the local tract structure. Most imaging studies have began using fractional anisotropy images in voxelwise statistical analyses so that it can localize the changes in the brain associated with development, disease and degeneration. Optimal analyses are however compromised especially through the application of standard registration algorithms (Douaud et al., 2011p 887). To this day there is no comprehensive solution for the problem of aligning fractional anisotropy images form many subjects in a manner that can produce acceptable conclusions to be made about subsequent voxelwise analysis. In addition, the arbitrary nature of the extent of spatial smoothing is yet to be resolved. However, tract-based spatial statistics (TBSS) has been found to have solutions to the above problems. The solutions are obtained through non linear registration that has been tuned carefully, to precede projection onto alignment invariant tract representation. The tract based spatial statistics is important because it makes sensitivity better, improves interpretability of analysis and objectivity of diffusion imaging studies with many subjects (Smith, Jenkinson, Johansen-Berg, Rueckert, Nichols & Mackay et al. 2006 p 1502).  Advantages and disadvantages The major advantages of the tract-based spatial statistics is that is a fully automated method of analysis that is easy to apply, it investigates the entire brain and does not need to have a localizing or pre-specifying features and regions of interest. (Wang et al 2011, 1580). TBSS is important because it reduces the difficulties associated with low resolution DTI data through Fractional Anisotropy value projection of specific subjects on to a single Fractional Anisotropy skeleton found in the main structures within the white matter. Linear as well as non linear alignment is used in the whole of this process and therefore they improve the degree of interpretation of the analysis of DTI data with many subjects (Seeley et al., 45). One disadvantage of the process is that this mapping cannot cope with high variability between individual brains. This happens more where there is cerebral atrophy as well as brain variability evident in aging and in a bigger way, neuro-generative disorders like Alzheimer’s disease. Other limitations to tract-based spatial statistics include issues resulting from inaccuracies in alignment and the failure to have a principled method of selecting smoothing extend. Approaches based on tractography have contradictory disadvantages and advantages (Smith et al 2006, p1488). They manage to overcome problems of alignment by operating in specific subject space tractography results and therefore they do not need pre-smoothing. These approaches however, are a kind of hindrance to the investigation of the entire brain and therefore they need the user to step in so that the tracts can be fully defined. Tract-based spatial statistics help to bring together the strong points in each of the approaches. The issues of alignment and smoothing are solved through automation and the entire brain is investigated without the particular identified tracts being pre-specified. Structural changes normally cause problems in image alignment to a predefined atlas especially if the atlas if this atlas originated from the scans of the brains of youthful healthy volunteers. The selection of an image for reference purposes may have an effect on the results and the way statistical comparisons for the cohorts are interpreted. Interpretation must consider the image to be used for reference purposes (Smith et al., 2007 p 500). Although tract-based spatial statistics (TBSS) tries to deal with registration error through a search in the neighborhood perpendicular to the skeleton of Fractional anisotropy for the voxel that has optimum FA, this step is incapable of compensating for bigger errors in registration which may appear in pathology such as the presence of atrophy in many neurogenerative diseases. This makes selection of reference and registration performance a critical issue. In some studies, researchers try to do away with the problem of misalignment through the modification of the registration in the TBSS pipeline (Smith, Johansen-Berg, Jenkinson, Rueckert, Nichols & Miller et al. 2007 p 499). Conclusion In conclusion, this essay has examined the Tract-based Spatial Statistics method of studying white matter in the brain when it is still in the process of development. The essay has discussed the nature of this method and its mode of use, the various studies done in the field TBSS, where it is used and its major advantages and disadvantages. TBSS has been applied in the study of the white matter in the brain and researchers have related it to the study of neurodegenerative disease disorders such as Alzheimer’s disease. Tract-based spatial statistics technique works automatically in the analysis of DTI where a technique based on voxel-wise comparison is used to critically examine associations between various subjects for example, the differences existing between varied groups. Researchers have undertaken various studies on TBSS in which they sought to discover different things. TBSS is used in hospitals and medical facilities for analyzing the white matter in the brain in relation to various diseases affecting the brain. TBSS is advantageous because it is automatic and easy to use. A major disadvantage is that mapping of the human brain cannot cope with high variability between individual brains and it has problem with errors in registration. Bibliography Douaud G, Jbabdi S, Behrens TEJ, Menke RA, Gass A, et al. (2011) DTI measures in crossing- fibre areas: Increased diffusion anisotropy reveals early white matter alteration in MCI and mild Alzheimer's disease. Neuroimage 55: 880–90 Keihaninejad S, Ryan NS, Malone IB, Modat M, Cash D, et al. (2012) The Importance of Group- Wise Registration in Tract Based Spatial Statistics Study of Neurodegeneration: A Simulation Study in Alzheimer's Disease. PLoS ONE 7(11): 45996. doi:10.1371/journal.pone.0045996 Seeley WW, Crawford RK, Zhou J, Miller BL, Greicius MD (2009) Neurodegenerative diseases target large-scale human brain networks. Neurone 62: 42–52.  Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, et al. (2006) Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 31: 1487–505.  Smith, S. M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T. E., & Mackay, C. E., et al. (2006). Tract-Based spatial statistics: voxelwise analysis multi-subject diffusion data. Neuroimage, 31(4), 1487-505. Smith, S. M., Johansen-Berg, H., Jenkinson, M., Rueckert, D., Nichols, T. E., & Miller, K. L., et al.(2007).Acquisition and voxelwise analysis of multi-subject diffusion data with tract- based spatial statistics. Nature Protocols, 2(3), 499-503. Smith, S., Jenkinson, M., Johansen-Berg, H., Ruekert, D., Nichols, T., Mackay, C., Watkins, K., Ciccarelli, O., Cader, M., Matthews, P. and Behrens, T. 2006. Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage, Volume 31, Issue 4. University of Oxford, UK. Pg.1487–1505. Wang Y, Gupta A, Liu Z, Zhang H, Escolar ML, et al. (2011) DTI registration in atlas based fiber analysis of infantile Krabbe disease. Neuroimage 55: 1577–86. Read More