Techniques that combine deterministic and statistical methods to solve inverse problems will be the focus of the mini-symposium. The speakers and audience may range from practitioners in medical imaging to more specialized mathematicians/statisticians working on applied inverse problems. This type of blended expertise is relevant to solving applied problems in imaging, particularly ones that are ill-posed in nature including electrical impedance tomography and optical tomography. In recent years, there has been a tremendous growth in devising new techniques both deterministic and statistical, such as model reduction using reduced basis method, sparsity, Bayesian inversion, Markov Chain Monte Carlo (MCMC) methods etc. The statistical approaches are becoming more popular due to the growth of computational power in the last several decades.

Damage Detection in Concrete Using Electrical Impedance Tomography: Deterministic and Statistical Perspectives

Taufiquar Khan (Clemson University)

Bayesian approach to optical flow in synthetic schlieren tomography

Aki Pulkkinen (University of Eastern Finland)

Electrical impedance tomography-based abdominal obesity estimation using deep learning

Minha Yoo (National Institute for Mathematical Sciences)

Imaging the solar interior

Thorsten Hohage (University of Goettingen )

Maximum-a-posteriori estimation with unknown regularisation parameters: combining deterministic and Bayesian approaches

Ana Fernandez Vidal (Heriot-Watt University)

Photoacoustic imaging using sparsity in curvelet frame

Bolin Pan (University College London)

Reconstructing a convex inclusion with one measurement of electrode data in the inverse conductivity problem

Minh Mach (University of Helsinki)

Recovery from model errors in magnetic particle imaging - approximation error modeling approach

Christina Brandt (University of Hamburg)

Deterministic methods for conductivity imaging for breast and skin cancer detection

Cristiana Sebu (University of Malta)

Total variation regularized Acousto-Electric Tomography with Neumann conditions

Bjørn Christian Skov Jensen (Technical University of Denmark)

Combining iterative and statistical inversion algorithms in imaging

Erkki Somersalo (Case Western Reserve University)

Regularization for Bayesian inverse problems using domain truncation and uncertainty quantification

Tan Bui-Thanh (The University of Texas at Austin)

Simulation of changes on optical coherence tomography data in healthy and in disease conditions

Aderito Araujo (University of Coimbra)

Gap Safe Screening Rules for Sparsity Enforcing Penalties

Ndiaye Eugene (Telecom ParisTech)

Organizers:

Taufiquar Khan (Clemson University)

Cristiana Sebu (University of Malta)

Keywords:

bayesian methods, computed tomography, deep learning, image enhancement, image reconstruction, integral equations for image analysis, inverse problems, nonlinear optimization, partial differential equation models, statistical inverse estimation methods, stochastic processes