Automotive
Victoria Mazo’s Research, From Two Layers of Graphene to a Stack of Medical Images
A physicist and computer vision researcher, Mazo has published on graphene and devised a way to find anatomical boundaries in a stack of medical images.
Victoria Mazo Hod Hasharon, MP, Israel
A single slice of a medical scan can leave an outline hard to distinguish. In a patent awarded in 2020, Victoria Mazo described an approach that lets a computer also consider the neighboring slices. Patterns on either side give the model more information about where an anatomical structure begins and ends in the image it is examining. Zebra Medical Vision filed the application in 2018, naming Mazo as the inventor.
Before that work, Mazo was publishing on a very different subject. A 2014 paper in Physical Review B lists her as first author with Bar-Ilan University’s physics department. The team used a mathematical model to study how electric and magnetic fields influence the behavior of electrons in bilayer graphene, a material just two atomic layers thick.
Mazo holds a doctorate in theoretical physics and describes working in computer vision at General Motors. Deep learning is another of her specialties. Her patent offers an unusually clear example of what those terms mean in practice: a researcher can change what a model sees by giving it more than one view of the same structure.
