Every year Inria@SiliconValley launches a call for Post-Doc Fellowships.
After a tough competition, we are please to congratulate Nina Miolane and Ziran Zhang who have been selected.
Nina Miolane Inria@SiliconValley Post-Doctoral Fellow
Nina Miolane will conduct her post-Doc within the GeomStats associate team between the ASCLEPIOS Inria Team (Sophia-Antipolis) and Susan Holmes’ team at Stanford University:
“Geometric Hierarchical Model of the Brain Anatomy”
Computational Medicine is a dynamic field experiencing an exponential growth. As an example, the cloud computing market for medical images is estimated at $400 millions for 2018 in the United States. This represents a 27% increase in ten years. Likewise, ”Medtech” – a term refering to technology for health care – is now a ”buzzword”: its web-based researches have increased by 20% since 2010. The transfer of medical imaging technology to industry and to the hospital is at the heart of this societal revolution.
The computational tools are especially important when it comes to neuroimaging. Brain images are produced every day in hospitals and show large anatomical variability. On the brain’s surface, for example, the distribution of folding patterns varies significantly from a healthy subject to another. Because of this complexity, the brain images are difficult to analyze. The clinician often extracts only visual qualitative information or quantitative information restricted to one dimension, for example a distance on a slice. Medicine needs computational tools that model the whole brain anatomy with its variability. This is required for a better understanding of the healthy and pathological brain variability. Associated to neuroimaging studies, it will improve our knowledge of diseases like Alzheimer or Parkinson for applications in hospitals. The postdoc research focuses on modeling the brain anatomical variability with a geometric hierarchical model.
.
Ziran Zhang Inria@SiliconValley Post-Doctoral Fellow
“SmartMarina”
The myth and fear of the performance reliability issue has hindered the pace of making wireless sensor networks (WSNs) a competitive alternative to their wired counterpart. Early wireless solutions were deemed as unreliable due to intermittent connectivity caused by single-frequency operation. Recent advances in low-power wireless technology, with a new architecture rooted in the “Time Synchronized Channel Hopping” (TSCH) link-layer technology, alleviated the adverse effect of multi-path fading, and improved the reliability of wireless links. However, there has been very little study done to identify the operational limits of such networks in the real-world applications by monitoring key network-performance-metrics.