Efficient inversion strategies for estimating optical properties with Monte Carlo radiative transport models
| Autor: | Samuel Powell, Simon R. Arridge, Callum M. Macdonald |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Paper
Optics and Photonics Computer science Monte Carlo method Biomedical Engineering ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION stochastic-gradient descent FOS: Physical sciences Iterative reconstruction optical tomography 01 natural sciences Bottleneck Data modeling 010309 optics Biomaterials radiative transport 0103 physical sciences medicine Image Processing Computer-Assisted Tomography Optical Physics - Biological Physics Optical tomography General Monte Carlo Tomographic reconstruction medicine.diagnostic_test Stochastic process Computational Physics (physics.comp-ph) Atomic and Molecular Physics and Optics Electronic Optical and Magnetic Materials Stochastic gradient descent machine learning Biological Physics (physics.bio-ph) Tomography X-Ray Computed Physics - Computational Physics Algorithm Monte Carlo Method Physics - Optics Optics (physics.optics) |
| Zdroj: | Journal of Biomedical Optics |
| DOI: | 10.48550/arxiv.2007.02601 |
| Popis: | Indirect imaging problems in biomedical optics generally require repeated evaluation of forward models of radiative transport, for which Monte Carlo is accurate yet computationally costly. We develop a novel approach to reduce this bottleneck which has significant implications for quantitative tomographic imaging in a variety of medical and industrial applications. Using Monte Carlo we compute a fully stochastic gradient of an objective function for a given imaging problem. Leveraging techniques from the machine learning community we then adaptively control the accuracy of this gradient throughout the iterative inversion scheme, in order to substantially reduce computational resources at each step. For example problems of Quantitative Photoacoustic Tomography and Ultrasound Modulated Optical Tomography, we demonstrate that solutions are attainable using a total computational expense that is comparable to (or less than) that which is required for a single high accuracy forward run of the same Monte Carlo model. This approach demonstrates significant computational savings when approaching the full non-linear inverse problem of optical property estimation using stochastic methods. Comment: 24 Pages, 11 figures |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|