Tomographic Imaging with Light and Sound Finnish Inverse Problems - PowerPoint PPT Presentation

Tomographic Imaging with Light and Sound Finnish Inverse Problems Summer School, Helsinki, June 3, 2019 Tanja Tarvainen, Computational Physics and Inverse Problems Group, Department of Applied Physics UEF // University of Eastern Finland

This is joint work with Aki Pulkkinen , Senior Researcher Jarkko Leskinen , postdoc Ultrasound physics and therapy Ultrasonic and optical instrumentation Acousto-optic interaction Eero Koponen , PhD student Aleksi Leino , postdoc Synthetic schlieren tomography Light transport simulations Niko Hänninen , PhD student Meghdoot Mozumder , postdoc Quantitative photoacoustic Diffuse optical tomography tomography Jenni Tick , PhD student Teemu Sahlström , PhD student Photoacoustic tomography Photoacoustic tomography 2 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Optical imaging – Why the interest? Early breast cancer imaging • In 1929 it was suggested that visible light could be used to detect breast lesions • M. Cutler: "Transillumination of the breast", 1929 •The method didn’t work 3 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Light has some properties which make it very appealing for biomedical studies • Capability of optical methods to provide information on the internal properties of tissues based on endogenous (e.g. haemoglobin) or exogenous (e.g. dyes) contrast • Instrumentation is relatively simple and low-cost • Non-ionizing Absorption coefficients of major chromophores in tissues (T. Näsi Multimodal applications of functional near-infrared spectroscopy, PhD thesis, Aalto University, 2013). 4 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• Various visible and near-infrared light based measurement and monitoring methods have been developed since 1930s • Pulse oximetry, NIR spectroscopy, ... Image from Wikimedia Commons, the free media repository. 5 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Diffuse optical tomography 1990- • Imaging through human body using visible and near-infrared light • Reconstructing distributions of light absorbers 6 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• Light is guided into the target from various direction • Amount of scattered and transmitted light is measured • An image is reconstructed from the measurements 7 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Diffusion approximation • Describes light propagation in highly scattering medium • Diffuse model • Easy to solve using numerical methods 8 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Image reconstruction in the framework of inverse problems • Solve the optical parameters which minimise functional using methods of computational inverse problems • The minimisation problem is solved using optimization methods • The governing (partial) differential equation needs to be solved using numerical methods • Both can be difficult tasks 9 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• S. Arridge and M. Schweiger: Inverse Methods for Optical Tomography, 1993 • S. Arridge et al: A finite element approach for modelling photon transport in tissue, 1993 • Etc. 10 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

However… • Diffuse optical tomography can only produce blurred (i.e. diffuse) images 11 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Imaging using coupled physics 2000- • Utilise coupled physics to combine the benefits of different imaging modalities • Photoacoustics, acousto-optics, ultrasound modulated diffuse optical tomography, thermoacoustics, etc. 12 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Photoacoustic imaging First experiments • Photoacoustic effect was first reported by Alexander Graham Bell in 1880 • Generated audio waves using chopped sunlight • A.G. Bell, On the production and Reproduction of Sound by Light, American Journal of Science, 20:305, 1880 • A.G. Bell, The Production of Sound by Radiant Energy, Science, 2(48):242-253, A.G. Bell, On the production and Reproduction of Sound by Light, American Journal of Science , 1881 20:305, 1880 13 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Photoacoustic effect 1. Tissue is illuminated by a short pulse (ns scale) of light 2. As light propagates within the tissue, it is absorbed by chromophores (light absorbing molecules) 14 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

3. The absorbed energy causes pressure rise 4. This pressure increase propagates though the tissue as an acoustic wave and can be detected on the surface of the tissue using ultrasound sensors 15 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Photoacoustic imaging • Reconstruct the initial pressure (or absorbed optical energy density) from the photoacoustic signal measured at the surface of the tissue 16 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• Photoacoustic imaging combines the benefits of optical and acoustic methods • Contrast though optical absorption – Tissue chromophores: oxygenated and deoxygenated haemoglobin, water, lipids, melanin – Contrast agents • Resolution by ultrasound – Low scattering in soft biological tissue J Laufer at al, Journal of Biomedical Optics, 2012 17 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Inverse problem in Bayesian framework • Model all parameters as random variables • The solution of the inverse problem (posterior probability density) is given by the Bayes’ formula 18 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• • 19 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• • 20 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

21 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

22 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

23 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

24 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

25 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Quantitative photoacoustic tomography • Estimate the distribution of the optical parameters from photoacoustic images 26 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

27 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Radiative transfer equation • Describes propagation of radiation in the presence of scattering particles 28 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• Is used to model light transport in astrophysics, atmosphere, oceanography, biomedical studies • Analytical solutions are limited to few geometries • Computationally expensive and challenging 29 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Optical inverse problem of QPAT • Estimate optical parameters x from absorbed optical energy density H • Forward model: radiative transfer equation or diffusion approximation • Maximum a posteriori estimate 30 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• 31 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

• Example from a simulation study 32 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

For more information • Computational physics and inverse problems research group: www.uef.fi/inverse • Biomedical Optical Imaging and Ultrasound Laboratory: www.uef.fi/opus • Open source Monte Carlo code and Matlab interface – ValoMC: https://inverselight.github.io/ValoMC/ 33 UEF // University of Eastern Finland Tomographic imaging with light and sound / Tanja Tarvainen 3.6.2019

Thank you!