Ilias Giannakopoulos

Ilias Giannakopoulos
NYU Langone Medical Center | NYUMC · Department of Radiology

PhD
Postdoctoral Fellow, CAI2R, CBI, Department of Radiology, NYU Grossman School of Medicine

About

26
Publications
1,969
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
110
Citations
Citations since 2016
26 Research Items
110 Citations
2016201720182019202020212022051015202530
2016201720182019202020212022051015202530
2016201720182019202020212022051015202530
2016201720182019202020212022051015202530
Introduction
I am a postdoctoral fellow at the CBI, Department of Radiology, NYU Grossman School of Medicine. I received my Ph.D. degree in Computational and Data Science and Engineering from Skoltech, CDISE, Moscow, Russia, in 2020, and my Diploma degree in electrical and computer engineering from AUTh, Greece, in 2016. I was an one-year (2019) visiting PhD student at MIT in the RLE.
Additional affiliations
November 2018 - November 2019
Massachusetts Institute of Technology
Position
  • Research Assistant
Description
  • Member of the Computational Prototyping Group
August 2016 - September 2020
Skolkovo Institute of Science and Technology
Position
  • Research Assistant
Description
  • Member of the Computational prototyping and imaging groups
Education
November 2018 - November 2019
Massachusetts Institute of Technology
Field of study
  • Electrical Engineering and Computer Science
August 2016 - September 2020
Skolkovo Institute of Science and Technology
Field of study
  • Computation Data Intensive Science & Engineering
October 2011 - May 2016
Aristotle University of Thessaloniki
Field of study
  • Electrical Engineering and Computer Science

Publications

Publications (26)
Article
Objective: In this paper, we introduce Global Maxwell Tomography (GMT), a novel, volumetric technique that estimates electric conductivity and permittivity by solving an inverse scattering problem based on magnetic resonance measurements. Methods: GMT relies on a fast volume integral equation solver, MARIE, for the forward path and a novel regul...
Article
Full-text available
We present a method of memory footprint reduction for FFT-based, electromagnetic (EM) volume integral equation (VIE) formulations. The arising Green’s function tensors have low multilinear rank, which allows Tucker decomposition to be employed for their compression, thereby greatly reducing the required memory storage for numerical simulations. Con...
Article
Full-text available
Objective: Global Maxwell Tomography (GMT) is a recently introduced volumetric technique for noninvasive estimation of electrical properties (EP) from magnetic resonance measurements. Previous work evaluated GMT using ideal radiofrequency (RF) excitations. The aim of this simulation study was to assess GMT performance with a realistic RF coil. Meth...
Article
Full-text available
In this work, we propose a method for the compression of the coupling matrix in volume-surface integral equation (VSIE) formulations. VSIE methods are used for electromagnetic analysis in magnetic resonance imaging (MRI) applications, for which the coupling matrix models the interactions between the coil and the body.We showed that these effects ca...
Preprint
Full-text available
Objective: We developed a hybrid volume surface integral equation (VSIE) method based on domain decomposition to perform fast and accurate magnetic resonance imaging (MRI) simulations that include both remote and local conductive elements. Methods: We separated the conductive surfaces present in MRI setups into two domains and optimized electromagn...
Article
High static field MR scanners can produce human tissue images of astounding clarity, but rely on high frequency electromagnetic radiation that generates complicated in-tissue field patterns that are patient-specific and potentially harmful. Many such scanners use parallel transmitters to better control field patterns, but then adjust the transmitte...
Conference Paper
Full-text available
Introduction: Electrical properties (EP), namely permittivity and electric conductivity, dictate the interactions between electromagnetic waves and biological tissue [1]. EP can be potential biomarkers for pathology characterization, such as cancer, and improve therapeutic modalities, such radiofrequency hyperthermia and ablation. MR-based electric...
Preprint
Full-text available
Electrical properties (EP), namely permittivity and electric conductivity, dictate the interactions between electromagnetic waves and biological tissue. EP can be potential biomarkers for pathology characterization, such as cancer, and improve therapeutic modalities, such radiofrequency hyperthermia and ablation. MR-based electrical properties tomo...
Conference Paper
Full-text available
At ultra­high­field magnetic resonance (MR) imaging (MRI), the inhomogeneity of the magnetic transmit + −and receive field maps (B1 , B1 ) encodes information about the distribution of tissue electrical properties (EP). Local EP reconstruction methods have been partially successful, although they are inherently limited by boundary artifacts and noi...
Article
Full-text available
Objective: We developed a hybrid volume surface integral equation (VSIE) method based on domain decomposition to perform fast and accurate magnetic resonance imaging (MRI) simulations that include both remote and local conductive elements. Methods: We separated the conductive surfaces present in MRI setups into two domains and optimized electrom...
Conference Paper
We propose a memory compression scheme for the coupling matrices appearing in volume-surface integral equation formulations. When there is some distance between the surface and the volumetric scatterers, the low-rank properties of the coupling matrix, allow us to reshape it into a set of four-dimensional tensors, which can be heavily compressed wit...
Poster
Global Maxwell Tomography (GMT) is a recently introduced technique that estimates tissue electrical properties from magnetic resonance measurements by solving an inverse scattering problem. In this work, we propose a new implementation of GMT that uses a Projected Newton method to minimize the cost function, instead of the quasi-Newton method emplo...
Poster
The volume-surface integral equation (VSIE) method is used for rapid and accurate simulations of electromagnetic fields in magnetic resonance imaging. For the case of a 7T array and a numerical head phantom, we constructed the VSIE coupling matrix that models the interactions between the coil and the scatterer. We reshaped the columns of the matrix...
Preprint
Full-text available
In this work, we propose a method for the compression of the coupling matrix in volume\hyp surface integral equation (VSIE) formulations. VSIE methods are used for electromagnetic analysis in magnetic resonance imaging (MRI) applications, for which the coupling matrix models the interactions between the coil and the body. We showed that these effec...
Article
A stable volume integral equation (VIE) solver based on polarization/magnetization currents is presented, for the accurate and efficient computation of the electromagnetic scattering from highly inhomogeneous and high contrast objects. We employ the Galerkin Method of Moments to discretize the formulation with discontinuous piecewise linear basis f...
Thesis
Full-text available
The scope of this doctoral dissertation is to study the interactions of electromagnetic (EM) waves and biological tissue in the presence of a strong magnetic field, and demonstrate the utility of novel methods via their application to simulations for the Magnetic Resonance Imaging (MRI) of realistic human head models. In problems related to MRI, th...
Poster
Electrical properties (EP) can be retrieved from magnetic resonance measurements. We employed numerical simulations to investigate the use of convolutional neural networks (CNN) as a tensor-to-tensor translation between transmit magnetic field pattern ( b1+ ) and EP distribution for simple tissue-mimicking phantoms. Given the volumetric nature of t...
Preprint
Full-text available
Objective: Global Maxwell Tomography (GMT) is a recently introduced volumetric technique for noninvasive estimation of electrical properties (EP) from magnetic resonance measurements. Previous work evaluated GMT using ideal radiofrequency (RF) excitations. The aim of this simulation study was to assess GMT performance with a realistic RF coil. Meth...
Conference Paper
Full-text available
We simulated a Global Maxwell Tomography experiment for the estimation of electrical properties in a numerical tissue-mimicking phantom using a decoupled 8 channel radiofrequency coil designed for 7 Tesla magnetic resonance scanners. The goal of this work was to investigate whether the orthogonality of the coil’s transmit fields (b1+ measurements)...
Conference Paper
Full-text available
In vivo access to electrical properties (EP) of biological tissues is a challenging problem that has tantalized scientists for decades. Non-invasive cross-sectional mapping of electrical conductivity and permittivity could be useful for tissue characterization and provide novel biomarkers for pathologies, including cancer. Furthermore, knowledge of...
Preprint
Full-text available
Objective: This paper proposes a stable volume integral equation (VIE) solver based on polarization/magnetization currents, for the accurate and efficient computation of the electromagnetic scattering from highly inhomogeneous and high contrast objects. Methods: We employ the Galerkin Method of Moments to discretize the formulation with discontinuo...
Conference Paper
Full-text available
We present an algorithm for the compression of the Green function tensors arising from fft-based volume integral equation formulations. The algorithm is based on an 3D adaptive cross approximation of the Tucker decomposition. We demonstrate that the reported method can lead to remarkable compression (x2000) for a typical example involving interacti...
Preprint
Full-text available
We present a method of memory footprint reduction for FFT-based, electromagnetic (EM) volume integral equation (VIE) formulations. The arising Green's function tensors have low multilinear rank, which allows Tucker decomposition to be employed for their compression, thereby greatly reducing the required memory storage for numerical simulations. Con...

Network

Cited By

Projects

Projects (3)
Project
Electrical properties (EP), namely permittivity and electric conductivity, dictate the interactions between electromagnetic waves and biological tissue. EP can be potential biomarkers for pathology characterization, such as cancer, and improve therapeutic modalities, such as radiofrequency hyperthermia and ablation. In this project, we aim to leverage the recent developments in physics-informed deep learning to solve the Helmholtz equation for the electrical property reconstruction of biological tissue in MRI frequencies.
Project
To develop methods for the memory footprint reduction of the electromagnetic volume and volume-surface integral equation formulations. The arising Green’s function tensors have low multilinear ranks, which allows Tucker and tensor train decompositions to be employed for their compression, thereby greatly reducing the required memory storage for numerical simulations. Consequently, the compressed components are able to fit inside a graphical processing unit on which highly parallelized computations can vastly accelerate the simulation time. The utility of these approaches is demonstrated via their application to problems of Magnetic Resonance Imaging.
Project
Global Maxwell Tomography (GMT) is a method for the electrical property reconstruction of human tissue. It is assumption-free, thus only employs measurable MR quantities for the formulation of an inverse problem. The goal is to minimize a cost-function between iteratively simulated MR measurements and experimental ones. For the solutions of the forward problem, the integral representation of Maxwell’s equations is employed. The method is 3D, and it is highly parallelizable to n independent problems, corresponding to each of the n excitations of the radiofrequency coil used. GMT is tested in simulation for 7 Tesla MRI frequencies, where the b1+ encodes additional anatomical information, in contrast to MR scanners with lower field strength, i.e., 3 and 1.5 Tesla.