PhD Position Available
The LPI is recruiting a PhD student to carry out his/her research in the diffusion MRI field. Questions should be directed to Professor Santiago Aja-Fernandez at firstname.lastname@example.org.
[Details (in Spanish)] [Overview (in Spanish)]
The position has been granted by the Spain Science Ministry and it is linked to the research grant entitled "Diffusion magnetic resonance imaging for precision medicine: from analysis to prediction. Application to migraine ". The PhD must be carried out in the University of Valladolid. The position is subject to an official application to the Ministry.
REQUIREMENTS: A Master's degree in computer science, signal processing, biomedical engineering or a related quantitative field (e.g., physics, electrical engineering) is required.
REFERRED QUALIFICATIONS: Familiarity with machine learning methods, statistical analysis, MRI fundamentals and/or image processing.
DEADLINE: 7th November 2019. Since the application must be done via an official web-based system, I recommend to contact us as soon as possible, in order to guide the candidate through all the paperwork.
THE CONTRACT: the Spanish Science Ministry offers an official research contract for 4 years, with a salary of 17,785 €/year (Valladolid is not an expensive place to live). The salary will increase up to 19,000 € after PhD, until the end of the 4th year. The University covers the Social Security of the student. The position offers 6,250€ extra to spend in research stays in other labs.
ADVISORS: The position is endorsed by Prof. Santiago Aja-Fernández (www.lpi.tel.uva.es/santi) and Prof. Antonio Tristan-Vega (www.lpi.tel.uva.es/node/525) from the Image Processing Lab (LPI), University of Valladolid.
SUMMARY OF THE PROJECT: The last two decades have proven the ability of Diffusion Magnetic Resonance Imaging (DMRI) to quantitatively describe multiple aspects of the white matter of the brain. This capacity has been successfully employed for medical research on numerous pathologies. However, beyond this capacity, DMRI could also have the potential to finally reach clinical practice, by providing single-subject prediction at several levels such as diagnosis, early detection and prediction of the evolution of pathological conditions (including response to treatment). In order to accomplish this goal, the integration of the DMRI information with other complementary modalities (structural MRI, fMRI, etc.) is required. In addition, suitable models are needed for the patterns of change of the brain structure from health to disease, or along its temporal evolution. This project aims precisely at the development of these elements in order to obtain a complete methodology that allows the use of DMRI in personalized medicine through single-subject predictions.
Accordingly, the first goal of this proposal is to define an optimal set of diffusion measures that provides a sufficiently detailed description of the diffusion in the brain while, at the same time, keeps the size of the acquired dataset to a minimum so that the acquisition process is suitable for realistic scenarios. Different DMRI models and techniques provide different scalar measures, to the point that there is a myriad of different possible diffusion models that can be applied and features that can be extracted from DMRI data. Many of them are incompatible from an acquisition perspective: some require specific sampling schemes, some need a great amount of samples with very long scanning times and some even demand dedicated hardware. It is therefore necessary to define a compatible set that is be able to capture different complementary aspects of the diffusion process. It will be also necessary to reformulate some of the existing measures.
The second goal of this work is to integrate the DMRI measures with other sources of information from brain imaging, especially other modalities of brain MRI that describe other features of brain anatomy, structure and function, in order to increase the amount of information available and, accordingly, its predictive power. Third, we seek to model the patterns of change that occur along the temporal evolution with development and aging, as well as the changes that differentiate a healthy from a pathological brain. This model will be used for single-subject prediction.
Finally, as a practical case, the new methodology will be applied to a personalized medicine approach to the study of migraine, through the collaboration with the Hospital Clínico Universitario de Valladolid. The structural differences in the brain between episodic and chronic migraine and the temporal evolution of the disease, in terms of response to treatment and the possible de-chronification, will be analyzed.