Computed Tomography using Differentiable Monte Carlo Simulation
Event details
| Date | 25.06.2024 |
| Hour | 15:00 › 17:00 |
| Speaker | Lovro Nuic |
| Location | |
| Category | Conferences - Seminars |
EDIC candidacy exam
Exam president: Prof. Mark Pauly
Thesis advisor: Prof. Wenzel Jakob
Co-examiner: Prof. Manuel Guizar-Sicairos
Abstract
X-ray Computed Tomography (CT) involves exposing
a sample to X-rays from multiple orientations and
measuring the fraction of radiation that passes through the
sample. Various reconstruction techniques have been developed to
leverage this data to reconstruct the sample’s internal structure.
Most reconstruction algorithms assume a linear nature of Xray
propagation through the object. This document will explore
different reconstruction algorithms and their underlying assumptions.
Finally, we will present our plans to develop a novel Xray
reconstruction methodology that embraces scattered X-rays
rather than attempting to eliminate them.
Background papers
A direct approach to estimating surfaces in tomographic data. Whitaker, R. T., & Elangovan, V. Medical Image Analysis, 2002. 6(3), 235-249.
https://dl.acm.org/doi/abs/10.1145/3528223.3530121 Neat: Neural adaptive tomography. Rückert, Darius, et al. ACM Transactions on Graphics (TOG), 2022, 41.4: 1-13.
X-ray computed tomography through scatter. Geva, Adam, et al. Proceedings of The European Conference on Computer Vision (ECCV). 2018.
Exam president: Prof. Mark Pauly
Thesis advisor: Prof. Wenzel Jakob
Co-examiner: Prof. Manuel Guizar-Sicairos
Abstract
X-ray Computed Tomography (CT) involves exposing
a sample to X-rays from multiple orientations and
measuring the fraction of radiation that passes through the
sample. Various reconstruction techniques have been developed to
leverage this data to reconstruct the sample’s internal structure.
Most reconstruction algorithms assume a linear nature of Xray
propagation through the object. This document will explore
different reconstruction algorithms and their underlying assumptions.
Finally, we will present our plans to develop a novel Xray
reconstruction methodology that embraces scattered X-rays
rather than attempting to eliminate them.
Background papers
A direct approach to estimating surfaces in tomographic data. Whitaker, R. T., & Elangovan, V. Medical Image Analysis, 2002. 6(3), 235-249.
https://dl.acm.org/doi/abs/10.1145/3528223.3530121 Neat: Neural adaptive tomography. Rückert, Darius, et al. ACM Transactions on Graphics (TOG), 2022, 41.4: 1-13.
X-ray computed tomography through scatter. Geva, Adam, et al. Proceedings of The European Conference on Computer Vision (ECCV). 2018.
Practical information
- General public
- Free