A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography

S. Zumbo; S. Mandija; T. Isernia; M. T. Bevacqua

A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography

S. Zumbo
, S. Mandija
, T. Isernia
, M. T. Bevacqua

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

Abstract

In this paper, we review and discuss a recent approach for Magnetic Resonance-based Electrical Properties Tomography that combines deep learning and physics-based calculations in an iteratively fashion. The performance of the approach is showed against a 2D simulated human brain models at 128 MHz, considering both noiseless and noisy data.

Original language	English
Title of host publication	9th Congress of the National Group of Bioengineering, GNB 2025 - Proceedings
Publisher	Patron Editore S.r.l.
ISBN (Electronic)	9788855584142
Publication status	Published - 2025
Event	9th Congress of the National Group of Bioengineering, GNB 2025 - Palermo, Italy Duration: 16 Jun 2025 → 18 Jun 2025

Publication series

Name	Convegno Nazionale di Bioingegneria
ISSN (Electronic)	2724-2129

Conference

Conference	9th Congress of the National Group of Bioengineering, GNB 2025
Country/Territory	Italy
City	Palermo
Period	16/06/25 → 18/06/25

Keywords

Convolutional neural network
Electrical properties tomography
inverse problems
magnetic resonance

Cite this

@inproceedings{f5512a1607ee4b0fa1977b6dc308b058,

title = "A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography",

abstract = "In this paper, we review and discuss a recent approach for Magnetic Resonance-based Electrical Properties Tomography that combines deep learning and physics-based calculations in an iteratively fashion. The performance of the approach is showed against a 2D simulated human brain models at 128 MHz, considering both noiseless and noisy data.",

keywords = "Convolutional neural network, Electrical properties tomography, inverse problems, magnetic resonance",

author = "S. Zumbo and S. Mandija and T. Isernia and Bevacqua, \{M. T.\}",

year = "2025",

language = "English",

series = "Convegno Nazionale di Bioingegneria",

publisher = "Patron Editore S.r.l.",

booktitle = "9th Congress of the National Group of Bioengineering, GNB 2025 - Proceedings",

}

A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography. / Zumbo, S.; Mandija, S.; Isernia, T. et al.
9th Congress of the National Group of Bioengineering, GNB 2025 - Proceedings. Patron Editore S.r.l., 2025. (Convegno Nazionale di Bioingegneria).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography

AU - Zumbo, S.

AU - Mandija, S.

AU - Isernia, T.

AU - Bevacqua, M. T.

PY - 2025

Y1 - 2025

N2 - In this paper, we review and discuss a recent approach for Magnetic Resonance-based Electrical Properties Tomography that combines deep learning and physics-based calculations in an iteratively fashion. The performance of the approach is showed against a 2D simulated human brain models at 128 MHz, considering both noiseless and noisy data.

AB - In this paper, we review and discuss a recent approach for Magnetic Resonance-based Electrical Properties Tomography that combines deep learning and physics-based calculations in an iteratively fashion. The performance of the approach is showed against a 2D simulated human brain models at 128 MHz, considering both noiseless and noisy data.

KW - Convolutional neural network

KW - Electrical properties tomography

KW - inverse problems

KW - magnetic resonance

UR - https://www.scopus.com/pages/publications/105033052124

M3 - Conference contribution

AN - SCOPUS:105033052124

T3 - Convegno Nazionale di Bioingegneria

BT - 9th Congress of the National Group of Bioengineering, GNB 2025 - Proceedings

PB - Patron Editore S.r.l.

T2 - 9th Congress of the National Group of Bioengineering, GNB 2025

Y2 - 16 June 2025 through 18 June 2025

ER -

A Physics-driven Learning Framework for MRI-Based Electrical Properties Tomography

Abstract

Publication series

Conference

Keywords

Other files and links

Fingerprint

Cite this