Jing Zhang

Ms. Jing Zhang

UBO, LaTIM, Brest, Bretagne, France
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 January 2026 in Physics in Medicine & Biology
Verified
Journal Article
Open access
DSDavid Sarrut
David Sarrut
NANicolas Arbor
Nicolas Arbor
TBThomas Baudier
Thomas Baudier
JBJulien Bert
Julien Bert
Konstantinos ChatzipapasKonstantinos Chatzipapas
Konstantinos Chatzipapas
NKNils Krah
Nils Krah
We present GATE version 10, a major evolution of the open-source Monte Carlo simulation application for medical physics, built on Geant4. This release marks a transformative evolution, featuring a modern Python-based user interface, enhanced multithreading and multiprocessing capabilities, the ability to be embedded as a library within other software, and a streamlined framework for collaborative development. In this Part 1 paper, we outline GATE’s position among other Monte Carlo codes, the core principles driving this evolution, and the robust development cycle employed. We also detail the new features and improvements. Part 2 will focus on the architectural innovations and technical challenges. By combining an open, collaborative framework with cutting-edge features, such a Monte Carlo platform supports a wide range of academic and industrial research, solidifying its role as a critical tool for innovation in medical physics.

25 January 2022 in Journal of Imaging
Verified
Journal Article
Open access
PDF file
Jing ZhangJing Zhang
Jing Zhang
Caroline PetitjeanCaroline Petitjean
Caroline Petitjean
SASamia Ainouz
Samia Ainouz
The fetus head circumference (HC) is a key biometric to monitor fetus growth during pregnancy, which is estimated from ultrasound (US) images. The standard approach to automatically measure the HC is to use a segmentation network to segment the skull, and then estimate the head contour length from the segmentation map via ellipse fitting, usually after post-processing. In this application, segmentation is just an intermediate step to the estimation of a parameter of interest. Another possibility is to estimate directly the HC with a regression network. Even if this type of segmentation-free approaches have been boosted with deep learning, it is not yet clear how well direct approach can compare to segmentation approaches, which are expected to be still more accurate. This observation motivates the present study, where we propose a fair, quantitative comparison of segmentation-based and segmentation-free (i.e., regression) approaches to estimate how far regression-based approaches stand from segmentation approaches. We experiment various convolutional neural networks (CNN) architectures and backbones for both segmentation and regression models and provide estimation results on the HC18 dataset, as well agreement analysis, to support our findings. We also investigate memory usage and computational efficiency to compare both types of approaches. The experimental results demonstrate that even if segmentation-based approaches deliver the most accurate results, regression CNN approaches are actually learning to find prominent features, leading to promising yet improvable HC estimation results.

14 January 2026 in Physics in Medicine & Biology
Verified
Journal Article
Open access
NKNils Krah
Nils Krah
NANicolas Arbor
Nicolas Arbor
TBThomas Baudier
Thomas Baudier
JBJulien Bert
Julien Bert
Konstantinos ChatzipapasKonstantinos Chatzipapas
Konstantinos Chatzipapas
DSDavid Sarrut
David Sarrut
Over the past years, we have developed GATE version 10, a major re-implementation of the long-standing Geant4-based Monte Carlo application for particle and radiation transport simulation in medical physics. This release introduces many new features and significant improvements, most notably a Python-based user interface replacing the legacy static input files. The new functionality of GATE version 10 is described in the part 1 companion paper (Sarrut et al 2025 arXiv:2507.09842). The development brought significant challenges. In this paper, we present the solutions that we have developed to overcome these challenges. In particular, we present a modular design that robustly manages the core components of a simulation: particle sources, geometry, physics processes, and data acquisition. The architecture consists of integrated C++ and Python codes. This framework allows for the precise, time-aware generation of primary particles, a critical requirement for accurately modeling positron emission tomography, radionuclide therapies, or prompt-gamma timing systems. We present how GATE 10 handles complex Geant4 physics settings while exposing a simple interface to the user. Furthermore, we describe the methodological solutions that facilitate the seamless integration of advanced physics models and variance reduction techniques. The architecture supports sophisticated scoring of physical quantities (such as Linear Energy Transfer and Relative Biological Effectiveness) and is designed for multithreaded execution. The new user interface allows researchers to script complex simulation workflows and directly couple external tools, such as artificial intelligence models for source generation or detector response. By detailing these architectural innovations, we demonstrate how GATE 10 provides a more powerful and flexible tool for research and innovation in medical physics. This paper is not intended to be a developer guide. Its purpose is to share with the research community in-depth explanations of our development effort that made the new GATE 10 possible.

21 January 2026 in Physics in Medicine & Biology
Verified
Journal Article
Closed access
Jing ZhangJing Zhang
Jing Zhang
ABAlexandre Bousse
Alexandre Bousse
CPChi-Hieu Pham
Chi-Hieu Pham
KSKuangyu Shi
Kuangyu Shi
JBJulien Bert
Julien Bert
Objective. Accurate and personalized radiation dose estimation is crucial for effective targeted radionuclide therapy (TRT). Deep learning (DL) holds promise for this purpose. However, current DL-based dosimetry methods require large-scale supervised data, which is scarce in clinical practice. Approach. To address this challenge, we propose exploring semi-supervised learning (SSL) framework that leverages readily available pre-therapy positron emission tomography (PET) data, where only a small subset requires dose labels, to predict radiation doses, thereby reducing the dependency on extensive labeled datasets. In this study, traditional classification-based SSL approaches were adapted and extended in regression task specifically designed for dose prediction. To facilitate comprehensive testing and validation, we developed a synthetic dataset that simulates PET images and dose calculation using Monte Carlo simulations. Main results. In the experiment, several regression-adapted SSL methods were compared and evaluated under varying proportions of labeled data in the training set. The overall mean absolute percentage error of dose prediction remained between 9% and 11% across different organs, which achieved comparable performance than fully supervised ones. Significance. The preliminary experimental results demonstrated that the proposed SSL methods yield promising outcomes for organ-level dose prediction, particularly in scenarios where clinical data are not available in sufficient quantities.

7 April 2026
Verified
Thesis
Closed access
Jing ZhangJing Zhang
Jing Zhang
Estimation des biomarqueurs à partir d'images médicales : approches basées sur la segmentation et sans segmentation: La segmentation est l'une des tâches les plus importantes dans l'analyse des images médicales. Depuis quelques années, les réseaux de neurones convolutifs (CNN) en constituent l'état de l'art. Dans ce contexte, nous allons nous focaliser sur les problématiques suivantes. Premièrement, la fonction de perte (loss) est une composante importante qui dirige l'apprentissage des CNN et décide de la relation entre les étiquettes cibles et les prédictions. Les fonctions de loss standard en particulier, telle que la loss de Dice, ont montré leurs limites. Deuxièmement, la segmentation est souvent la première étape pour ensuite estimer les paramètres (également appelés biomarqueurs) de l'image. Ces biomarqueurs sont utilisés pour établir un diagnostic et un suivi des patients. Une estimation précise des biomarqueurs est donc capital. Cependant, des erreurs sont susceptibles de se produire lors de l'étape intermédiaire de segmentation. Récemment, les techniques d'apprentissage profond ont ouvert la voie à l'estimation directe des biomarqueurs à partir des images, sans segmentation ou extraction de caractéristique adhoc. La recherche sur ce sujet en est encore à ses débuts. Pour répondre à ces questions, cette thèse propose les contributions suivantes, résumées en trois points : tout d'abord, nous proposons une nouvelle fonction de perte, basée sur le coefficient Kappa, qui a la capacité de prendre en compte tous les pixels de l'image, y compris le vrai négatif, contrairement à la perte standard de Dice. Nous illustrons sa valeur ajoutée sur un jeu de données public d'images de lésions cutanées. Deuxièmement, nous contribuons à la prédiction directe de biomarqueurs sans segmentation afin de fournir une solution d'analyse raisonnable et efficace pour les applications cliniques. Nous proposons plusieurs architectures de CNN de régression, qui apprennent directement à estimer les paramètres d'intérêt sans recourir à la segmentation. Un cas d'application est la prédiction de la circonférence de la tête du fœtus à partir d'images échographiques : nous comparons segmentation et régression avec un protocole expérimental judicieux. De ce fait, nous avons pu montrer des résultats prometteurs pour la régression, même si des améliorations restent possibles. Un autre cas est la prédiction des volumes de la structure cardiaque à partir d'images de résonance magnétique tridimensionnelles, dans lequel une méthode de prédiction multi-objectifs est réalisée. Troisièmement, nous étudions l'interprétabilité des modèles de régression, en étendant les techniques standard de cartes de saillance aux CNN de régression, qualitativement et quantitativement. Nous avons pu montrer que, sur la plupart des images, le CNN de régression apprend réellement à identifier la zone cible