Speaker
Description
Our research demonstrates the feasibility of utilizing compact X-ray setup coupled with advanced computational techniques for accurate and efficient breast cancer diagnostics. We present study of Small-Angle X-ray Scattering (SAXS) in human tissues through Geant4 Monte Carlo simulations. Scattering events from 1 mm thick tissue is recorded with sensitive detector placed at 20 mm (WAXS) and 160 mm (SAXS) from the sample. Materials composed of four components are used to artificially generate tissues which are then used to calculate 2D scattering images. With pyFAI azimutal integration 2D image is transformed to the 1D scattering vector in reciprocal space and appended in dataset. We then use obtained dataset to train a classifier capable of distinguishing between different material compositions ( data labels) by providing as input 1d scattering data (features vector).
| Abstract publication | I agree that the abstract will be published on the web site |
|---|
