Keywords: Imaging biomarkers, multi-compartment models, statistical analysis
Thesis topic
Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system affecting more than one million persons in North America or in Europe. It is characterized by widespread inflammation, focal demyelination, and a variable degree of axonal loss. Today, conventional MRI is widely used for MS diagnosis, patient follow-up and monitoring of therapies. MRI abnormalities in MS are seen with high sensitivity although they are non-specific. Non-conventional quantitative MRI (e.g. diffusion tensor imaging, relaxometry) has proven to provide useful information on the extent of diffuse tissue injuries (outside lesions) at all stages of the disease. Consequently, it allows to better understand the physiopathology of the disease. However, DTI is mostly used through scalar measures that may lose information contained in the diffusion tensors at the voxel level. Moreover, DTI is known to be insufficient to characterize crossing fiber pathways. Diffusion compartment imaging (DCI) [9,10], modeling several fibers per voxel, enables a much finer characterization. Also, sub-voxel integration of new quantitative biomarkers derived from DCI and relaxometry [1,4,6] is a real challenge relative to the various spatial resolutions and artifact sensitivities of these methods, but also with regard to the state-of-the-art models that derive parameters from these data.
The overall goal of this PhD is to solve the clinical-MRI paradox by introducing a sophisticated new tissue compartment imaging capability enabling us to measure disease burden and severity in neural circuits. The clinical-radiological paradox in MS is the observation that patient symptoms do not reflect the disease burden and progression seen with MRI from the current imaging technologies [2,3,5]. In this PhD, we intend to perform an integration of multi-compartment neural circuits imaging biomarkers that will introduce a novel, more specific disease evolution score. Novel imaging strategies [7] will be proposed to combine the improved characterization of neural circuits provided by DCI (through microstructure descriptors) and MRI quantitative relaxation times specific to these circuits. Such innovative descriptors will be computed for major neural circuits extracted from DCI tractography [8] to propose new, more specific measures of disease burden and lesion severity. This will result in increased correlation between disease burden/lesion severity and functional specialization of the cortex, which directly relates to patient symptoms.
The PhD candidate will develop a mathematical framework that enables registration and computation of atlases of brain microstructure models. These new methods will in turn enable novel statistical analyses of the brain microstructure, specifically with a method called Fascicle-Based Spatial Statistics (FBSS), which will automatically detect microstructural abnormalities along fascicles rather than on a voxel basis, as achieved by DTI and tract-based spatial statistics (TBSS) that currently constitute the state-of-the-art. In addition to this statistical framework, contrary to many approaches that rely on per-voxel comparison of data, the PhD candidate will integrate disease severity indexes along extracted neural circuits to compute a per-circuit, per-compartment evaluation of disease burden. We will for example consider as disease burden scores the computation of the average of DCI parameters in the compartments involved in the neural circuit, together with the myelin water fraction (a critical tissue affected by MS) along the tracts extracted from novel MR relaxometry acquisitions. Such disease burden per-tract scalar indexes could be computed for controls and patients. We will then be able to compute scores characterizing for a given patient and neural circuit how the patient score belongs to the same distribution as the one of controls.
Location
This PhD will take place at Inria/IRISA, UMR CNRS 6074, in the VisAGeS U746 research team. The work will be conducted in close link with our MRI experimental platform (http://www.neurinfo.org). This PhD will be performed in the context of the joint international lab between Inria and Harvard Medical School (http://team.inria.fr/barbant) in close collaboration between Visages and the Computational Radiology Laboratory at the Children’s Hospital, Boston.
Requirements: C++, Matlab, strong knowledge on statistics and applied mathematics: signal and/or image processing, some knowledge of MRI acquisition techniques will be appreciated as well as fluent English.
References
1. Akhondi-Asl, A., O. Afacan, R. V. Mulkern and S. K. Warfield (2014). "T2-Relaxometry for Myelin Water Fraction Extraction Using Wald Distribution and Extended Phase Graph." MICCAI.
2. R. Bakshi, A. J. Thompson, M. A. Rocca, D. Pelletier, V. Dousset, F. Barkhof, M. Inglese, et al. (2008). "MRI in multiple sclerosis: current status and future prospects." Lancet Neurol 7(7): 615-625.
3. Barkhof, F. (2002a). "The clinico-radiological paradox in multiple sclerosis revisited." Current opinion in neurology 15(3): 239-245.
4. Cao, F., O. Commowick, E. Bannier and C. Barillot (2014). "Simultaneous Estimation of T1, T2 and B1 Maps From Relaxometry MR Sequences." MICCAI Workshop on Intelligent Imaging Linking MR Acquisition and Processing: 16- 23.
5. Hackmack, K., M. Weygandt, J. Wuerfel, C. F. Pfueller, J. Bellmann-Strobl, F. Paul and J. D. Haynes (2012). "Can we overcome the 'clinico-radiological paradox' in multiple sclerosis?" J Neurol 259(10): 2151-2160
6. S.H. Kolind, B. Mädler, S. Fischer, D. K. Li and A. L. MacKay (2009). "Myelin water imaging: implementation and development at 3.0T and comparison to 1.5T measurements." MRM 62(1): 106-115
7. Scherrer, B. and S. K. Warfield (2012a). "Parametric representation of multiple white matter fascicles from cube and sphere diffusion MRI." PLoS One 7(11).
8. Stamm, A., O. Commowick, C. Barillot and P. Pérez (2013). "Adaptive multi-modal particle filtering for probabilistic white matter tractography." Information Processing in Medical Imaging. 7917: 1-12.
9. Stamm, A., P. Pérez and C. Barillot (2012). "A new multi-fiber model for low angular resolution diffusion MRI." IEEE International Symposium on Biomedical Imaging: 936-939.
10. Stamm, A., B. Scherrer, O. Commowick, C. Barillot and S. K. Warfield (2014). "Fast and robust detection of the optimal number of fascicles in diffusion images using model averaging theory." ISMRM. 22.
RTP Slot Gacor memang menjadi pilihan menarik bagi banyak pemain karena peluang menangnya yang lebih besar. Dengan persentase RTP Slot Tertinggi, pemain bisa merasa lebih percaya diri saat bermain dan berharap pada peluang yang lebih baik. Saat bermain di RTP Slot Online, sensasi adrenalin saat melihat peluang kemenangan yang tinggi menjadi pengalaman tersendiri. Slot RTP Tertinggi menawarkan stabilitas dan frekuensi kemenangan yang sangat dinantikan banyak pemain.