Bio sketch
After graduating in Bioengineering and Life Science Technology at the Ecole Polytechnique Fédérale de Lausanne (EPFL), I joined the Translational Machine Learning Lab to pursue a PhD in a collaborative setting between EPFL, CHUV and Siemens Healthineers.
I was given the invaluable opportunity to evolve in a very dynamic research environment benefitting from strong collaborations between Siemens Healthineers and expert radiologists and researchers from various institutions, including the University Hospital of Lausanne, the Inselspital of Bern and the Charles University Hospital of Prague.
During my PhD I wrote 4 first author journal papers, contributed to 7 publications as a co-author and published 28 conference abstracts.
My PhD thesis
My PhD thesis aimed at developing new quantitative radiological biomarkers for neurological diseases, and particularly multiple sclerosis, to improve patient diagnosis and prognosis. In the process of defining such advanced biomarkers, particular attention was given to enabling a successful translation of the developed methods into clinical practice. In this context, we proposed for instance the use of population-based atlases instead of advanced diffusion weighted imaging – which requires longer acquisition times – alongside with the use of quantitative MRI (qMRI) maps instead of conventional weighted MRI contrasts to reduce inter- scan variability.
In the attempt of defining clinically relevant biomarkers to bridge the clinico-radiological gap in MS, we investigated three hypotheses: first, diffuse microstructural alterations of normal-appearing tissue, detected with qMRI, improves the characterization of disability; second, the lesion location plays an important role in brain function; third, different types of lesions (i.e., active, inactive, chronically active lesions) cause different degrees of brain damage severity and thus observed disability.
Despite being a powerful tool to reduce inter-scan variability, qMRI techniques are still not widely used in clinical practice. In a side project, I also investigated domain adaptation methods to harmonize T1-weighted MRI protocols between different institutions, to reduce the inter-site variability of segmentation algorithms.
Awards
| May 2021 virtual | Quantitative MR study group award, ISMRM Best Quantitative MR Clinical Translation Abstract: “T1 abnormalities in atlas-based white matter tracts: reducing the clinico-radiological paradox in multiple sclerosis using qMRI”. |
| May 2021 virtual | ISMRM magna cum laude merit award Prize awarded the top 15% abstracts presented during the ISMRM & SMRT meeting: ”The sensitivity of classical and deep image similarity metrics to MR acquisition parameters”. |
| Oct. 2019 Lausanne, Switzerland | Annaheim Matille Award, Fondation Marguerite Rewards a high-level master project devoted to bringing together life sciences and information technology. |
| Oct. 2019 Lausanne, Switzerland | Computational Sciences Award, IBM Research Rewards either a master project or a doctoral thesis in order to promote excellent research in modelling and simulation in different fields of engineering and science. |
| Oct. 2019 Lausanne, Switzerland | Best Master Thesis in Bioegineering, EPFL Section of Life Sciences Engineering |