Valorisez vos réalisations phares sur People@HES-SO Plus d'infos
PEOPLE@HES-SO – Annuaire et Répertoire des compétences
PEOPLE@HES-SO – Annuaire et Répertoire des compétences

PEOPLE@HES-SO
Annuaire et Répertoire des compétences

Aide
language
  • fr
  • en
  • de
  • fr
  • en
  • de
  • SWITCH edu-ID
  • Administration
ID
« Retour
Schmid Jérôme

Schmid Jérôme

Professeur HES ordinaire

Compétences principales

Imagerie médicale

Artificial Intelligence (AI)

Traitement d'images

Digital image processing

Signal Processing (Images/Video)

Computer Vision

Machine Learning

  • Contact

  • Enseignement

  • Recherche

  • Publications

  • Conférences

Contrat principal

Professeur HES ordinaire

Haute école de santé - Genève
Avenue de Champel 47, 1206 Genève, CH
HEDS-GE
Domaine
Santé
Filière principale
Technique en radiologie médicale

Le professeur Jérôme Schmid travaille depuis de nombreuses années dans le domaine du traitement et de l'analyse des images médicales, en se concentrant sur la segmentation et le recalage. Il s'est spécialisé dans les modèles déformables à base physique appliqués à la modélisation du système musculo-squelettique humain. Le professeur Schmid a également exploré la combinaison des modèles déformables avec l'apprentissage automatique, ainsi que l'utilisation de l'apprentissage profond pour le diagnostic assisté par ordinateur, comme la reconnaissance des fractures osseuses dans les radiographies du poignet, la détection de la maladie de Parkinson dans l'imagerie SPECT ou la détection des lésions mammaires dans l'IRM dynamique. L'application de l'intelligence artificielle à des fins pédagogiques a également été explorée par son équipe lors de la conception du prototype de formation AIRx pour simuler la génération réaliste de radiographies.

Depuis 2011, le professeur Schmid et son équipe à la HES-SO ont acquis une solide expertise en matière de diagnostic assisté par ordinateur et de chirurgie guidée par l'image dans le cadre de plusieurs projets de recherche financés par les secteurs public et privé, notamment le Fonds national suisse de la recherche scientifique, lnnosuisse et diverses fondations suisses.

MSc HES-SO/UNIL en Sciences de la santé - HES-SO Master
  • Intelligence artificielle et médecine assistée par ordinateur
  • Acquisition et post-traitement des images radiologiques
  • Systèmes informatisés de santé
  • Mathématiques appliquées à l'imagerie radiologique
BSc HES-SO en Technique en radiologie médicale - Haute école de santé - Genève
  • Propriétés de l'image numérique
  • Reconstruction, traitement et visualisation d'images radiologiques
  • Etude de cas multimodalités
  • Méthodologie de recherche
  • Intelligence artificielle et médecine assistée par ordinateur
Technologies et simulation 3D en orthopédie et médecine du sport - Faculté de médecine de l'Université de Genève
  • Reconstruction, traitement et visualisation d'images radiologiques

En cours

DeepDAT: Investigating techniques of artificial intelligence for the diagnosis of Parkinson’s disease with Dopamine SPECT imaging

Rôle: Requérant(e) principal(e)

Financement: Partenaire du secteur privé

Description du projet :

La prévalence mondiale de la maladie de Parkinson (MP) a plus que doublé au cours de la dernière génération. Les coûts associés à la MP représentent une charge économique substantielle pour la société suisse et, bien que la maladie réagisse très bien au traitement médical, il est de plus en plus évident qu'une intervention précoce peut contribuer à améliorer la qualité de vie des patients. La technique d'imagerie SPECT du transporteur de la dopamine (DAT) a prouvé sa valeur clinique en améliorant sensiblement la confiance des médecins dans leur diagnostic de la maladie, mais elle dépend du niveau d'expertise des lecteurs. Si l'intelligence artificielle (IA) peut apporter une aide précieuse au diagnostic de la MP, ses performances dépendent fortement de la disponibilité et de la qualité des données d'entraînement. Notre objectif est d'étudier les performances de l'IA pour détecter la MP dans des ensembles de données d'imagerie hétérogènes, collectés avec différents équipements et à différents endroits. Une attention particulière sera accordée à la conception d'outils d'IA interprétables pour soutenir la prise de décision clinique.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme

Partenaires académiques: Prof. Garibotto Valentina, Hôpitaux Universitaires de Genève

Durée du projet: 01.02.2022 - 01.06.2025

Statut: En cours

SUBREAM: Smart And Ultrafast Breast MR Imaging For Cancer Detection

Rôle: Requérant(e) principal(e)

Financement: Recherche suisse contre le cancer

Description du projet :

Le cancer du sein est le cancer le plus fréquent et la principale cause de décès par cancer chez les femmes, avec une moyenne de 1'400 décès par an en Suisse. Bien que l'imagerie par résonance magnétique (IRM) présente la meilleure sensibilité, son utilisation pour le dépistage des patients présentant des risques faibles à modérés est remise en question en raison de son coût élevé et de la disponibilité limitée des équipements. La philosophie du projet SUBREAM est de "faire mieux avec moins", en tirant parti de séquences IRM intelligentes et des avancées en intelligence artificielle. En effet, la pratique clinique actuelle de l'imagerie par résonance magnétique du sein repose encore fortement sur la caractérisation morphologique des lésions dans les images pré-contraste à haute résolution et l'imagerie de diffusion, et exploite des descripteurs élaborés manuellement à partir de la cinétique de contraste mesurée sur plusieurs séries DCE pondérées en T1 - ce qui conduit à des examens longs et sophistiqués qui peuvent encore caractériser à tort les lésions et être responsables de biopsies inutiles. Le projet SUBREAM étudiera un nouveau protocole IRM du sein visant à raccourcir de manière significative le temps d'acquisition et à améliorer les performances diagnostiques - en combinant les progrès récents de l'IA et des séquences IRM ultrarapides du sein.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Hyacinthe Jean-Noël , Mataj-Lokaj Belinda , Chênes Christophe

Partenaires académiques: Dr. Kinkel Karen, Réseau hospitalier neuchâtelois; Prof. Lovis Christian, Hôpitaux Universitaires de Genève / Unige

Partenaires professionnels: Dr. Djema Dahila, Clinique des Grangettes

Durée du projet: 01.01.2022 - 30.06.2025

Montant global du projet: 227'750 CHF

Url du site du projet: https://gap.swisscancer.ch/publicportal#/details/KFS-5460-08-2021

Statut: En cours

Terminés

Intelligence artificielle et transformation de la profession TRM
AGP

Rôle: Co-requérant(s)

Financement: HES-SO Rectorat; Santé; VD-HESAV

Description du projet : L'intelligence artificielle (IA) se développe et s'implante dans tous les secteurs, y compris ceux de la vie quotidienne, et suscite des mutations majeures dans le monde professionnel. C'est notamment le cas du domaine de la santé, et de l'imagerie médicale en particulier. Notre projet propose une étude de terrain exploratoire, dont l'objectif est d'analyser de quelle manière les technicien·ne·s en radiologie médicale utilisent l'IA dans leur activité professionnelle et quelles transformations de leur profession cela implique. Une enquête ethnographique sera menée dans un groupe d'imagerie médicale privé en Suisse romande.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Rey Séverine , Al-Musibli Azal

Durée du projet: 01.11.2021 - 31.08.2022

Montant global du projet: 70'551 CHF

Statut: Terminé

AIRx: Investigating techniques of artificial intelligence in the teaching of radiography

Rôle: Requérant(e) principal(e)

Financement: Fondation Hasler

Description du projet :

The “AIRx” project aims to introduce and exploit techniques of artificial intelligence (AI) in the teaching of radiography (Rx). The primary targeted audience are radiographer students, i.e. the future healthcare professionals mainly responsible of the acquisition and processing of medical images. The main objective is to deliver a radiography simulator using AI to detect human positioning and generate artificial yet realistic X-ray images. Indeed, some inadequacies in current teaching approaches are mainly explained by safety and ethical reasons: students perform real X-ray acquisitions on inanimate objects only, while they practice positioning exercises on other students without acquiring X-ray images; conversely, internships expose students to live situations – but these situations cannot be controlled like in teaching practice. Another objective is to promote the teaching of AI to health professionals while reinforicing their ICT knowledge, which is becoming a serious priority at both national and international levels. We expect innovative results in the field of ICT and in particular AI, and a practical impact on the learning experience of X-ray imaging.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Chênes Christophe

Durée du projet: 01.11.2019 - 31.12.2020

Montant global du projet: 49'854 CHF

Statut: Terminé

High-temperature tolerant X-ray imaging detector for medical applications in harsh environments

Rôle: Co-requérant(s)

Requérant(e)s: Thiran Jean-Philipe, EPFL

Financement: Innosuisse - Swiss Innovation Agency

Description du projet :

The overwhelming majority of medical X-ray systems provided in emerging markets are still film-based, incurring very high operating costs and yielding poor image quality. Modern digital X-ray systems are not spread in these countries due to their high complexity and fragility, but also because their X-ray detectors are not working properly at high temperatures. This project aims to develop a medical x-ray detector able to provide high-quality images in very harsh environments. This detector combines state-of-the-art CMOS APS sensors in a novel hardware architecture for real-time image processing.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Chênes Christophe

Partenaires académiques: Thiran Jean-Philipe, EPFL; Verdun Francis, CHUV, Institut de radiophysique appliquée (IRA)

Partenaires professionnels: Blanchard Hubert, Pristem SA

Durée du projet: 01.03.2018 - 30.07.2020

Montant global du projet: 408'985 CHF

Statut: Terminé

Development and Validation of A Multi-modal Image-based Model Generation Method for Non-invasive Dynamic Assessment of Femoroacetabular Impingement (FAI)

Rôle: Co-requérant(s)

Requérant(e)s: Zheng Guoyan, Université de Berne, ISTB

Financement: Fonds National Suisse (FNS)

Description du projet :

ENGLISH:

Femoroacetabular impingements (FAI) is defined as a painful dynamic abutment between the hip socket and the femoral head. It is a major cause of primary hip osteoarthritis and one of the key risk factors that may lead to early cartilage and labral damage in young adults. The therapy of FAI involves the surgical resection of the impinging areas of the acetabulum and the proximal femur. Based on a strong research experience on FAI, we know that its diagnosis is difficult and that the current gold standard is problematic as it requires the direct intraoperative visualization of the impingement conflict. A less invasive approach relies on a radiological analysis of standard 2D radiographs and MR images - a task however not automated and whose quality depends on the physician’s expertise. Virtual 3D impingement simulations provide a more objective and accurate way to study FAI but most approaches rely on CT scans, which expose patients to high doses of radiation. MR images were used as a non-invasive alternative to build 3D models for FAI analysis but the resulting approaches were often not fully automatic or required conditions sometimes incompatible with clinical setups (e.g., MR images with large field-of-view (FOV))

This project will develop an efficient method to generate 3D anatomical models using Computed Tomography (CT)-free imaging protocols that are used in clinical routine in order to support computer-assisted diagnosis and surgical planning of FAI. We will devise a fully automatic approach based on multi-modal images combining 2D X-ray radiograph with 3D Magnetic Resonance (MR) Images acquired with small FOV. In order to achieve this goal, we are aiming to develop A) a fully automatic, machine learning based approach for segmenting 3D MRI data acquired with small FOV; B) a disease-specific, articulated statistical shape model (DS-aSSM) based 2D-3D reconstruction technique to generate 3D patient-specific models from 2D X-rays; and C) a robust multi-modal data fusion method to fuse models generated from A) and B).Scientific and Social Impact of the Research ProjectAlthough there are many risk factors contributing to hip osteoarthritis (HOA), it is now generally accepted that more subtle bony abnormalities associated with FAI and/or developmental dysplasia (DD) contribute substantially to HOA. Various studies have suggested that computer-assisted 3D modeling techniques can help clinicians to better understand the dynamic and static factors of FAI - via patient-specific pre-operative planning and intra-operative assessment. Capitalizing on standard FAI diagnosis imaging protocol, i.e., conventional X-rays and MRI, to reduce radiation and to avoid unnecessary disruption to the standard clinical workflow, the proposed CT-free 3D anatomical model generation approach will facilitate a future wide-spread access of computational simulation and virtual surgical planning techniques for patients with FAI and address a key social challenge of our society.

FRENCH:

Le projet vise à développer des approches assistées par ordinateur pour créer des modèles personnalisées de la hanche dans le but d’étudier et traiter les conflits fémoro-acétabulaires.

L’originalité du projet réside dans l’utilisation d’acquisitions radiologiques très peu invasives (radiographies et IRM) et dans la création de méthodes informatisées robustes et totalement automatisées.

Les conflits fémoro-acétabulaires sont une des causes majeures de coxarthrose primaire et sont à l’origine de lésions du cartilage et du labrum chez les jeunes adultes. Il est estimé que 10 à 15% de la population adulte serait atteinte par une des formes de ces conflits. Afin de restituer au patient mobilité et confort, le traitement chirurgical est souvent nécessaire. Celui-ci nécessite alors une planification précise qui se base notamment sur la localisation des conflits sur les surfaces articulaires.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Chênes Christophe

Partenaires académiques: Zheng Guoyan, Université de Berne, ISTB; Tannast Moritz, Université de Berne, Inselspital

Durée du projet: 01.07.2018 - 30.04.2019

Montant global du projet: 174'463 CHF

Statut: Terminé

GlobalDiagnostiX: diagnostic approprié pour les hôpitaux des pays en voie de développement

Rôle: Co-requérant(s)

Requérant(e)s: Schönenberger Klaus, Ecole polytechnique fédérale de Lausanne (EPFL): EssentialMed

Financement: HES-SO

Description du projet :

La radiologie et l’échographie sont des outils essentiels de la médecine. Pourtant, ces instruments médicaux d’importance vitale sont inaccessibles pour plus de deux tiers de la population mondiale. La cause principale de ce problème d’accès est l’inadéquation entre la technologie existante et le contexte des pays pauvres.

Le projet GlobalDiagnostiX constitue une approche innovante et ambitieuse visant à contribuer à la solution  d'un problème majeur des systèmes de santé des pays pauvres : le développement et déploiement d’une technologie appropriée en imagerie médicale pour améliorer l’accès à ce moyen diagnostic clé au niveau des petits hôpitaux.

Chapeauté par la fondation EssentialMed du Centre de Coopération et Développement (CODEV) de l’Ecole Polytechnique Fédérale de Lausanne (EPFL), le projet GlobalDiagnostiX vise à développer un système d'imagerie diagnostique adapté au contexte des pays pauvres ayant un coût total de cycle de vie inférieur à $50'000 (y-compris les coûts d’achat et d’exploitation pendant 10 ans).

Actuellement, une importante coalition d’instituts de la Haute Ecole Spécialisée de Suisse Occidentale (HES-SO), de l’EPFL et de l’institut Paul Scherrer (PSI) travaille sur le développement d’un prototype. La filière Technique en Radiologie Médicale (TRM) de la Haute école de Santé de Genève (HEdS) participe à son développement, recherche des solutions innovantes et s’assure que le prototype réponde aux besoins des futurs utilisateurs.

Afin de pallier au manque de formation et assurer un soutien en continu en faveur de l’apprentissage Utilisateur sur place, la filière TRM de la HEdS travaille sur le développement d’une plateforme de formation et de soutien technologique, notamment en s’appuyant sur des solutions de type « télémédecine ».

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Fleury Eric , Chênes Christophe

Partenaires académiques: Schönenberger Klaus, Ecole polytechnique fédérale de Lausanne (EPFL): EssentialMed; EPFL, DESL - LTS5 – LSR – CODEV-EssentialTech; HES-SO, Valais, heigVD IESE / COMATEC, hepia inSTI, Edana, écal; Paul Scherrer Instritute (PSI), LSB; University Research Center on Ennergy for Health Care (CURES), University Research Center on Ennergy for Health Care (CURES); Centre hospitalier universitaire vaudois, Department of Medical Radiology; Swiss Tropical and Public Health Institute (Swiss TPH), SSCIH

Partenaires professionnels: Hôpital Universitaire de Yaoundé – CHU

Durée du projet: 01.02.2012 - 31.01.2019

Url du site du projet: http://www.globaldiagnostix.org/

Statut: Terminé

MyPlanner: implementing advanced processing of clinical 2D X-ray imaging into optimal individualized planning and intraoper-ative assistance for total joint arthroplasty

Rôle: Requérant(e) principal(e)

Financement: Innosuisse - Swiss Innovation Agency

Description du projet :

 

Ce projet vise à améliorer la planification chirurgicale d'arthroplastie totale en développant des approches qui réduisent la dose ionisante délivrée par l'imagerie radiologique tout en offrant un outil de planification précis aux chirurgiens. Le projet se concentre donc sur les protocoles d'acquisition radiologiques ainsi que sur le traitement informatisée de l'image radiologique.

L’utilisation du scanner (tomodensitométrie) dans l’arthroplastie totale a révolutionné la planification chirurgicale en offrant plus de précision et de reproductibilité grâce à la modélisation tridimensionnelle et personnalisée des articulations du patient. Cependant l’utilisation de cette modalité d’imagerie s’est accompagnée d’une augmentation de la dose ionisante délivrée au patient.

Ce projet vise à rebasculer de la tomodensitométrie à la radiographie conventionnelle en fournissant une solution qui fournirait une qualité de reconstruction anatomique comparable et qui s’intégrerait de façon harmonieuse dans le workflow clinique. Les techniques développées offriront un support aux chirurgiens orthopédiques à tous les niveaux: planification préopératoire, assistance peropératoire par guides de coupe et enfin analyse postopératoire du placement des implants.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Chênes Christophe

Partenaires professionnels: Bernardoni Massimiliano, Medacta International SA

Durée du projet: 01.09.2014 - 31.12.2018

Montant global du projet: 145'360 CHF

Statut: Terminé

MyHip: Patient-Specific Pre-operative Planning and Intra-operative Surgical Guidance for Total Hip Arthroplasty

Rôle: Requérant(e) principal(e)

Financement: Commission pour la technologie et l'innovation (CTI)

Description du projet :

ENGLISH:

MyHip aims at creating a complete solution for the planning and execution of total hip arthroplasty (THA). By considering preoperatively the kinematics of the prosthetic hip and pelvis tilting, accurately modeled from the patient morphology, a dynamic planning for THA can be defined. This novel planning provides operating parameters which aim to minimize the onset of articular conflicts and (sub-)luxations. Furthermore, the creation of intra-operative guidance tools for the faithful reproduction of the surgical planning strengthens this dynamic planning for THA.

In this project, Artanim was responsible for the development of the dynamic simulation module of the prosthetic hip driven by motion capture data.

FRENCH:

L’arthroplastie totale de la hanche consiste à remplacer l’articulation défaillante et douloureuse par divers composants prothétiques. Cette chirurgie généralement efficace conduit parfois à des mécanismes fonctionnels indésirables qui impliquent une révision précoce de la prothèse.
Ces effets découlent en général d’une planification mal adaptée et caractérisée par un mauvais choix et positionnement des implants. La planification conventionnelle se base essentiellement sur une analyse de clichés radiographiques, adoptant ainsi une approche « statique » qui ignore à tort la dynamique de l’articulation prothétique.


Le projet MyHip se base sur l’hypothèse qu’une planification, qui considère de façon personnalisée la dynamique de la hanche prothétique, apportera au patient un meilleur confort et une efficacité mécanique sur le long terme.
En se basant sur l’imagerie médicale et la capture de mouvement optique, la morphologie du patient, son inclinaison pelvienne ainsi que sa mobilité sont modélisés et utilisés dans une simulation informatique. Cette assistance par ordinateur permet de détecter tous conflits articulaires et de corriger ainsi le choix et la pose des implants.


Le projet contribue aussi au développement de guides de coupes, qui permettent de reproduire de façon plus fiable et plus précise le planning chirurgical.

Equipe de recherche au sein de la HES-SO: Schmid Jérôme , Chênes Christophe

Partenaires académiques: Charbonnier Caecilia, Fondation Artanim; Christofilopoulos Panayiotis, Hôpitaux Universitaires de Genève

Partenaires professionnels: Bernardoni Massimiliano, Medacta International SA

Durée du projet: 01.02.2012 - 31.01.2014

Montant global du projet: 269'750 CHF

Url du site du projet: https://www.hesge.ch/heds/recherche-developpement/projets-recherche/myhip

Statut: Terminé

2025

Multimodal deep learning fusion of ultrafast-DCE MRI and clinical information for breast lesion classification
Article scientifique ArODES

Belinda Lokaj, Valentin Durand de Gevigney, Amal-Dahila Djema, Jamil Zaghir, Jean-Philippe Goldman, Mina Bjelogrlic, Hugues Turbé, Karen Kinkel, Christian Lovis, Jerome Schmid

Computers in biology and medicine,  2025, 188, 109721

Lien vers la publication

Résumé:

Background : Breast cancer is the most common cancer worldwide, and magnetic resonance imaging (MRI) constitutes a very sensitive technique for invasive cancer detection. When reviewing breast MRI examination, clinical radiologists rely on multimodal information, composed of imaging data but also information not present in the images such as clinical information. Most machine learning (ML) approaches are not well suited for multimodal data. However, attention-based architectures, such as Transformers, are flexible and therefore good candidates for integrating multimodal data. Purpose : The aim of this study was to develop and evaluate a novel multimodal deep learning (DL) model combining ultrafast dynamic contrast-enhanced (UF-DCE) MRI images, lesion characteristics and clinical information for breast lesion classification. Materials and methods : From 2019 to 2023, UF-DCE breast images and radiology reports of 240 patients were retrospectively collected from a single clinical center and annotated. Imaging data were constituted of volumes of interest (VOI) extracted around segmented lesions. Non-imaging data were constituted of both clinical (categorical) and geometrical (scalar) data. Clinical data were extracted from annotated reports and were associated to their corresponding lesions. We compared the diagnostic performances of traditional ML methods for non-imaging data, an image model based on the DL architecture, and a novel Transformer-based architecture, the Multimodal Sieve Transformer with Vision Transformer encoder (MMST-V). Results : The final dataset included 987 lesions (280 benign, 121 malignant lesions, and 586 benign lymph nodes) and 1081 reports. For classification with non-imaging data, scalar data had a greater influence on performances of lesion classification (Area under the receiver operating characteristic curve (AUROC) = 0.875 ± 0.042) than categorical data (AUROC = 0.680 ± 0.060). MMST-V achieved better performances (AUROC = 0.928 ± 0.027) than classification based on non-imaging data (AUROC = 0.900 ± 0.045), and imaging data only (AUROC = 0.863 ± 0.025). Conclusion : The proposed MMST-V is an adaptative approach that can consider redundant information provided by multimodal information. It demonstrated better performances than unimodal methods. Results highlight that the combination of clinical patient data and detailed lesion information as additional clinical knowledge enhances the diagnostic performances of UF-DCE breast MRI.

Characterization of fluid in facial sinuses on post-mortem CT in case of death by drowning
Article scientifique ArODES

Lucia Fernandes Mendes, Leonor Pedreira Lago, Coraline Egger, Jerome Schmid

International journal of legal medicine,  2025, to be published

Lien vers la publication

Résumé:

Recent statistics show that drowning deaths are a reality in Switzerland. Although drowning remains a diagnosis by exclusion for forensic pathologists, post-mortem multidetector computed tomography (PMCT) is a real complementary resource to establish the cause of death. The aim of this study was to determine whether the sinuses’ fluids visualized in post-mortem MDCT in case of drowning have specific characteristics that can be related to the submersion process and provide an additional element to the diagnosis of death by drowning. A few studies exist on the subject, but none of them compared putrefied bodies and drowned cases. A balanced dataset was retrospectively collected of 108 cases of natural death, putrefied bodies and drowned cases at the University Center of Legal Medicine in Geneva. For each paranasal cavity, the fluid and sinus were segmented in the slice with largest liquid quantity to derive liquid-to-sinus surface ratio (LR), and mean density (MD) in Hounsfield units (HU) of the fluid when present. For all sinuses, the MD was significantly different between putrefied and drowned groups. The average LR was statistically different for frontal and maxillary sinuses. Using cut-off values as Youden indices from ROC curves, promising specificities (Sp) and sensibilities (Se) were obtained, using single (e.g., frontal sinus: LR cut-off = 0.15: Sp = 76%, Se = 68%; MD cut-off = 44,55HU: Sp = 93%, Se = 64%, or maxillary sinus: LR cut-off = 0.14: Sp = 56%, Se = 86%; MD cut-off = 34,91HU, Sp = 65%, Se = 85%) or all (logistic regression: Sp = 80%, Se = 92.6%) measurements. This study identified potential leads for discrimination of drowning cases from natural deaths and putrefied bodies.

Learning MRI with ImmeRgaMe :
Article scientifique ArODES
exploring the pedagogical potential of an innovative serious game for radiographer training

Marie-Anaïs Petit, Joël Piguet, Belinda Lokaj, Jerome Schmid, Céline Gaignot

Radiography,  2025, 31, 3, 102921

Lien vers la publication

Résumé:

Introduction : This study aimed to assess the potential of ImmeRgaMe, an innovative 360° immersive serious game, to enhance motivation and engagement in MRI education for first-year radiographer students. Methods : The platform was developed as part of an educational innovation initiative and designed to complement traditional teaching methods. It integrates storytelling, progression mechanisms, and interactive tools to teach MRI safety protocols, physical principles, and clinical methods. Beta testing involved 42 participants (students and educators) who completed standardized surveys to evaluate gameplay, usability, and learning outcomes. The impact of microlearning videos embedded in the game was also assessed through pre- and post-test quizzes with 63 first-year students. Results : The beta version received positive feedback, with over 90 % of participants rating the gameplay and integrated tools as satisfactory. Students demonstrated improved performance in quizzes after using the microlearning videos, with an average score increase of 19 % across tests. Survey results indicated that 97 % of respondents found the videos helpful for understanding course content, and 86 % believed the game could support their progress during MRI internships. Conclusion : ImmeRgaMe effectively fosters student motivation and engagement, bridging the gap between theoretical and practical knowledge in MRI training. While promising, further refinement and broader testing are needed to evaluate its impact on knowledge retention and skills development, as well as its applicability to other imaging modalities. Implications for practice : The implementation of serious games like ImmeRgaMe in radiography education could reshape and modernize traditional teaching methods. By adapting this approach to other imaging modalities, educators could offer immersive and interactive learning experiences, fostering self-regulated learning and aligning training with the complex demands of clinical practice.

2024

Efficient clinical information extraction from breast radiology reports in French
Chapitre de livre ArODES

Jamil Zaghir, Belinda Lokaj, Karen Kinkel, Amal-Dahila Djema, Hugues Turbé, Mina Bjelogrlic, Valentin Durand de Gevigney, Jerome Schmid, Christian Lovis, Jean-Philippe Goldman

Dans Andrikopoulou, Elisavet, Arvanitis, Theodoros N., Benis, Arriel, Demiris, George, Gallos, Parisis, Hasman, Arie, Mantas, John, Marschollek, Michael, Ognjanovic, Ivana, Saranto, Kaija, Zoulias, Emmanouil, Studies in health technology and informatics  (5 p.). 2024,  Amsterdam : IOS Press Pays-Bas : IOS Press

Lien vers la publication

Résumé:

Radiology reports contain crucial patient information, in addition to images, that can be automatically extracted for secondary uses such as clinical support and research for diagnosis. We tested several classifiers to classify 1,218 breast MRI reports in French from two Swiss clinical centers. Logistic regression performed better for both internal (accuracy > 0.95 and macro-F1 > 0.86) and external data (accuracy > 0.81 and macro-F1 > 0.41). Automating this task will facilitate efficient extraction of targeted clinical parameters and provide a good basis for future annotation processes through automatic pre-annotation.

Barriers and facilitators of artificial intelligence conception and implementation for breast imaging diagnosis in clinical practice :
Article scientifique ArODES
a scoping review

Belinda Lokaj, Marie-Thérèse Pugliese, Karen Kinkel, Christian Lovis, Jerome Schmid

European radiology,  2024, 34, 3, 2096-2109

Lien vers la publication

Résumé:

Objective : Although artificial intelligence (AI) has demonstrated promise in enhancing breast cancer diagnosis, the implementation of AI algorithms in clinical practice encounters various barriers. This scoping review aims to identify these barriers and facilitators to highlight key considerations for developing and implementing AI solutions in breast cancer imaging. Method : A literature search was conducted from 2012 to 2022 in six databases (PubMed, Web of Science, CINHAL, Embase, IEEE, and ArXiv). The articles were included if some barriers and/or facilitators in the conception or implementation of AI in breast clinical imaging were described. We excluded research only focusing on performance, or with data not acquired in a clinical radiology setup and not involving real patients. Results : A total of 107 articles were included. We identified six major barriers related to data (B1), black box and trust (B2), algorithms and conception (B3), evaluation and validation (B4), legal, ethical, and economic issues (B5), and education (B6), and five major facilitators covering data (F1), clinical impact (F2), algorithms and conception (F3), evaluation and validation (F4), and education (F5). Conclusion : This scoping review highlighted the need to carefully design, deploy, and evaluate AI solutions in clinical practice, involving all stakeholders to yield improvement in healthcare.

2023

A novel image augmentation based on statistical shape and intensity models :
Article scientifique ArODES
application to the segmentation of hip bones from CT images

Jerome Schmid, Lazhari Assassi, Christophe Chênes

European radiology experimental,  August 2023, vol. 7, article 39

Lien vers la publication

Résumé:

Background : The collection and annotation of medical images are hindered by data scarcity, privacy, and ethical reasons or limited resources, negatively affecting deep learning approaches. Data augmentation is often used to mitigate this problem, by generating synthetic images from training sets to improve the efficiency and generalization of deep learning models. Methods : We propose the novel use of statistical shape and intensity models (SSIM) to generate augmented images with variety in both shape and intensity of imaged structures and surroundings. The SSIM uses segmentations from training images to create co-registered tetrahedral meshes of the structures and to efficiently encode image intensity in their interior with Bernstein polynomials. In the context of segmentation of hip joint (pathological) bones from retrospective computed tomography images of 232 patients, we compared the impact of SSIM-based and basic augmentations on the performance of a U-Net model. Results : In a fivefold cross-validation, the SSIM augmentation improved segmentation robustness and accuracy. In particular, the combination of basic and SSIM augmentation outperformed trained models not using any augmentation, or relying exclusively on a simple form of augmentation, achieving Dice similarity coefficient and Hausdorff distance of 0.95 [0.93–0.96] and 6.16 [4.90–8.08] mm (median [25th–75th percentiles]), comparable to previous work on pathological hip segmentation. Conclusions : We proposed a novel augmentation varying both the shape and appearance of structures in generated images. Tested on bone segmentation, our approach is generalizable to other structures or tasks such as classification, as long as SSIM can be built from training data.

2022

X-ray imaging detector for radiological applications adapted to the context and requirements of low- and middle-income countries
Article scientifique ArODES

Mario Andrés Chavarria, Matthias Huser, Sebastien Blanc, Pascal Monnin, Jerome Schmid, Christophe Chênes, Lazhari Assassi, Hubert Blanchard, Romain Sahli, Jean-Philippe Thiran, René Salathé, Klaus Schönenberger

Review of scientific instruments,  2022, vol. 93, no. 3, article 034102

Lien vers la publication

Résumé:

This paper describes the development of a novel medical x-ray imaging system adapted to the needs and constraints of low- and middle-income countries. The developed system is based on an indirect conversion chain: a scintillator plate produces visible light when excited by the x rays, and then, a calibrated multi-camera architecture converts the visible light from the scintillator into a set of digital images. The partial images are then unwarped, enhanced, and stitched through parallel field programmable gate array processing units and specialized software. All the detector components were carefully selected focusing on optimizing the system’s image quality, robustness, cost-effectiveness, and capability to work in harsh tropical environments. With this aim, different customized and commercial components were characterized. The resulting detector can generate high quality medical diagnostic images with detective quantum efficiency levels up to 60% (@2.34 μGy), even under harsh environments, i.e., 60 °C and 98% humidity.

2021

A new 2D-3D registration gold-standard dataset for the hip joint based on uncertainty modeling
Article scientifique ArODES

Fabio D'Isidoro, Christophe Chênes, Stephen J. Ferguson, Jerome Schmid

Medical physics,  2021, vol. 48, no. 10, pp. 5991-6006

Lien vers la publication

Résumé:

Purpose : Estimation of the accuracy of 2D-3D registration is paramount for a correct evaluation of its outcome in both research and clinical studies. Publicly available datasets with standardized evaluation methodology are necessary for validation and comparison of 2D-3D registration techniques. Given the large use of 2D-3D registration in biomechanics, we introduced the first gold standard validation dataset for computed tomography (CT)-to-x-ray registration of the hip joint, based on fluoroscopic images with large rotation angles. As the ground truth computed with fiducial markers is affected by localization errors in the image datasets, we proposed a new methodology based on uncertainty propagation to estimate the accuracy of a gold standard dataset. Methods : The gold standard dataset included a 3D CT scan of a female hip phantom and 19 2D fluoroscopic images acquired at different views and voltages. The ground truth transformations were estimated based on the corresponding pairs of extracted 2D and 3D fiducial locations. These were assumed to be corrupted by Gaussian noise, without any restrictions of isotropy. We devised the multiple projective points criterion (MPPC) that jointly optimizes the transformations and the noisy 3D fiducial locations for all views. The accuracy of the transformations obtained with the MPPC was assessed in both synthetic and real experiments using different formulations of the target registration error (TRE), including a novel formulation of the TRE (uTRE) derived from the uncertainty analysis of the MPPC. Results : The proposed MPPC method was statistically more accurate compared to the validation methods for 2D-3D registration that did not optimize the 3D fiducial positions or wrongly assumed the isotropy of the noise. The reported results were comparable to previous published works of gold standard datasets. However, a formulation of the TRE commonly found in these gold standard datasets was found to significantly miscalculate the true TRE computed in synthetic experiments with known ground truths. In contrast, the uncertainty-based uTRE was statistically closer to the true TRE. Conclusions : We proposed a new gold standard dataset for the validation of CT-to-X-ray registration of the hip joint. The gold standard transformations were derived from a novel method modeling the uncertainty in extracted 2D and 3D fiducials. Results showed that considering possible noise anisotropy and including corrupted 3D fiducials in the optimization resulted in improved accuracy of the gold standard. A new uncertainty-based formulation of the TRE also appeared as a good alternative to the unknown true TRE that has been replaced in previous works by an alternative TRE not fully reflecting the gold standard accuracy.

Revisiting Contour-Driven and Knowledge-Based Deformable Models: Application to 2D-3D Proximal Femur Reconstruction from X-ray Images
Article scientifique

Chênes Christophe, Schmid Jérôme

International Conference on Medical Image Computing and Computer-Assisted Intervention - MICCAI 2021, Lecture notes in Computer Sciences (LNCS), Part VI, 2021 , vol.  12906, pp.  451-460

Lien vers la publication

2020

Deformable multisurface segmentation of the spine for orthopedic surgery planning and simulation
Article scientifique ArODES

Rabia Haq, Jerome Schmid, Roderick Borgie, Joshua Cates, Michel A. Audette

Journal of medical imaging,  2020, vol. 7, no. 1, article 015002

Lien vers la publication

Résumé:

Purpose: We describe a shape-aware multisurface simplex deformable model for the segmentation of healthy as well as pathological lumbar spine in medical image data. Approach: This model provides an accurate and robust segmentation scheme for the identification of intervertebral disc pathologies to enable the minimally supervised planning and patient-specific simulation of spine surgery, in a manner that combines multisurface and shape statistics-based variants of the deformable simplex model. Statistical shape variation within the dataset has been captured by application of principal component analysis and incorporated during the segmentation process to refine results. In the case where shape statistics hinder detection of the pathological region, user assistance is allowed to disable the prior shape influence during deformation. Results: Results demonstrate validation against user-assisted expert segmentation, showing excellent boundary agreement and prevention of spatial overlap between neighboring surfaces. This section also plots the characteristics of the statistical shape model, such as compactness, generalizability and specificity, as a function of the number of modes used to represent the family of shapes. Final results demonstrate a proof-of-concept deformation application based on the open-source surgery simulation Simulation Open Framework Architecture toolkit. Conclusions: To summarize, we present a deformable multisurface model that embeds a shape statistics force, with applications to surgery planning and simulation.

2019

An automated segmentation framework for nasal computational fluid dynamics analysis in computed tomography
Article scientifique ArODES

Robin Huang, Anthony Nedanoski, David F. Fletcher, Narinder Singh, Jerome Schmid, Paul M. Young, Nicholas Stow, Lei Bi, Daniela Traini, Eugene Wong, Craig L. Philips, Ronald R. Grunstein, Jinman Kim

Computers in biology and medicine,  2019, vol. 115, 103505

Lien vers la publication

Résumé:

The use of computational fluid dynamics (CFD) to model and predict surgical outcomes in the nasal cavity is becoming increasingly popular. Despite a number of well-known nasal segmentation methods being available, there is currently a lack of an automated, CFD targeted segmentation framework to reliably compute accurate patient-specific nasal models. This paper demonstrates the potential of a robust nasal cavity segmentation framework to automatically segment and produce nasal models for CFD. The framework was evaluated on a clinical dataset of 30 head Computer Tomography (CT) scans, and the outputs of the segmented nasal models were further compared with ground truth models in CFD simulations on pressure drop and particle deposition efficiency. The developed framework achieved a segmentation accuracy of 90.9 DSC, and an average distance error of 0.3 mm. Preliminary CFD simulations revealed similar outcomes between using ground truth and segmented models. Additional analysis still needs to be conducted to verify the accuracy of using segmented models for CFD purposes.

Segmentation of the proximal femur in radial MR scans using a random forest classifier and deformable model registration
Article scientifique ArODES

Dimitrios Damapoulos, Till Dominic Lerch, Florian Schmaranzer, Moritz Tannast, Christophe Chênes, Guoyan Zheng, Jerome Schmid

International journal of computer assisted radiology and surgery,  2019, vol. 14, no. 3, pp. 545-561

Lien vers la publication

Résumé:

Background : Radial 2D MRI scans of the hip are routinely used for the diagnosis of the cam type of femoroacetabular impingement (FAI) and of avascular necrosis (AVN) of the femoral head, both considered causes of hip joint osteoarthritis in young and active patients. A method for automated and accurate segmentation of the proximal femur from radial MRI scans could be very useful in both clinical routine and biomechanical studies. However, to our knowledge, no such method has been published before. Purpose : The aims of this study are the development of a system for the segmentation of the proximal femur from radial MRI scans and the reconstruction of its 3D model that can be used for diagnosis and planning of hip-preserving surgery. Methods : The proposed system relies on: (a) a random forest classifier and (b) the registration of a 3D template mesh of the femur to the radial slices based on a physically based deformable model. The input to the system are the radial slices and the manually specified positions of three landmarks. Our dataset consists of the radial MRI scans of 25 patients symptomatic of FAI or AVN and accompanying manual segmentation of the femur, treated as the ground truth. Results : The achieved segmentation of the proximal femur has an average Dice similarity coefficient (DSC) of 96.37 ± 1.55%, an average symmetric mean absolute distance (SMAD) of 0.94 ± 0.39 mm and an average Hausdorff distance of 2.37 ± 1.14 mm. In the femoral head subregion, the average SMAD is 0.64 ± 0.18 mm and the average Hausdorff distance is 1.41 ± 0.56 mm. Conclusions : We validated a semiautomated method for the segmentation of the proximal femur from radial MR scans. A 3D model of the proximal femur is also reconstructed, which can be used for the planning of hip-preserving surgery.

Towards a deformable multi-surface approach to ligamentous spine models for predictive simulation-based scoliosis surgery planning
Article scientifique

Michel Audette, Schmid Jérôme, Craig Goodmurphy, Michael Polanco, Sebastian Bawab, Austin Tapp, Sheldon St-Clair

CSI 2018: Computational Methods and Clinical Applications for Spine Imaging, 2019 , pp.  90-102

Lien vers la publication

2015

MyHip :
Article scientifique ArODES
supporting planning and surgical guidance for a better total hip arthroplasty

Jerome Schmid, Christophe Chênes, Sylvain Chagué, Pierre Hoffmeyer, Panayiotis Christofilopoulos, Massimiliano Bernardoni, Caecilia Charbonnier

International journal of computer assisted radiology and surgery,  2015, vol. 10, pp. 1547-1556

Lien vers la publication

Résumé:

Purpose : Total hip arthroplasty (THA) aims to restore patient mobility by providing a pain-free and stable artificial joint. A successful THA depends on the planning and its execution during surgery. Both tasks rely on the experience of the surgeon to understand the complex biomechanical behavior of the hip. We investigate the hypothesis that a computer-assisted solution for THA effectively supports the preparation and execution of the planning. Methods : We devised MyHip as a computer-assisted framework for THA. The framework provides pre-operative planning based on medical imaging and optical motion capture to optimally select and position the implant. The planning considers the morphology and range of motion of the patient’s hip to reduce the risk of impingements and joint instability. The framework also provides intra-operative support based on patient-specific surgical guides. We performed a post-operative analysis on three patients who underwent THA. Based on post-operative radiological images, we reconstructed a patient-specific model of the prosthetic hip to compare planned and effective positioning of the implants. Results : When the guides were used, we measured non-significant variations of planned executions such as bone cutting. Moreover, patients’ hip motions were acquired and used in a dynamic simulation of the prosthetic hip. Conflicts prone to implant failure, such as impingements or subluxations, were not detected. Conclusions : The results show that MyHip provides a promising computer assistance for THA. The results of the dynamic simulation highlighted the quality of the surgery and especially of its planning. The planning was properly executed since non-significant variations were detected during the radiological analysis.

Analysis of hip range of motion in everyday life :
Article scientifique ArODES
a pilot study

Caecilia Charbonnier, Sylvain Chagué, Jerome Schmid, Frank C. Kolo, Massimiliano Bernardoni, Panayiotis Christofilopoulos

HIP international,  January-February 2015, vol. 25, no. 1, pp. 82-90

Lien vers la publication

Résumé:

Patients undergoing total hip arthroplasty are increasingly younger and have a higher demand concerning hip range of motion. To date, there is no clear consensus as to the amplitude of the “normal hip” in everyday life. It is also unknown if the physical examination is an accurate test for setting the values of true hip motion. The purpose of this study was: 1) to precisely determine the necessary hip joint mobility for everyday tasks in young active subjects to be used in computer simulations of prosthetic models in order to evaluate impingement and instability during their practice; 2) to assess the accuracy of passive hip range of motion measurements during clinical examination. A total of 4 healthy volunteers underwent Magnetic Resonance Imaging and 2 motion capture experiments. During experiment 1, routine activities were recorded and applied to prosthetic hip 3D models including nine cup configurations. During experiment 2, a clinical examination was performed, while the motion of the subjects was simultaneously captured. Important hip flexion (mean range 95°-107°) was measured during daily activities that could expose the prosthetic hip to impingement and instability. The error made by the clinicians during physical examination varied in the range of ±10°, except for flexion and abduction where the error was higher. This study provides useful information for the surgical planning to help restore hip mobility and stability, when dealing with young active patients. The physical examination seems to be a precise method for determining passive hip motion, if care is taken to stabilise the pelvis during hip flexion and abduction.

Segmentation of X-ray Images by 3D-2D Registration Based on Multibody Physics
Article scientifique

Schmid Jérôme, Chênes Christophe

ACCV 2014 - Lecture Notes in Computer Science, 2015 , vol.  9004, pp.  674-687

Lien vers la publication

MyHip: supporting planning and surgical guidance for a better total hip arthroplasty
Article scientifique

Schmid Jérôme, Chênes Christophe, Sylvain Chagué, Pierre Hoffmeyer, Panayiotis Christofilopoulos, Massimiliano Bernardoni, Caecilia Charbonnier

International Journal of Computer Assisted Radiology and Surgery, 2015 , vol.  10, pp.  1547-1556

Lien vers la publication

2014

Post-operative Kinematics Assessment in Patients after Total Hip Arthroplasty: A Pilot Study
Article scientifique

Caecilia Charbonnier, Sylvain Chagué, Chênes Christophe, Pierre Hoffmeyer, Panayiotis Christofilopoulos, Massimiliano Bernardoni, Schmid Jérôme

Swiss Medical Weekly, 2014 , vol.  144

Analysis of hip range of motion in everyday life: A pilot study
Article scientifique

Caecilia Charbonnier, Sylvain Chagué, Schmid Jérôme, Frank C. Kolo, Massimiliano Bernardoni, Panayiotis Christofilopoulos

HIP International, 2014 , vol.  25, no  1, pp.  82-90

Lien vers la publication

2013

Hip Range of Motion in Everyday Life
Article scientifique

Panayiotis Christofilopoulos, Sylvain Chagué, Schmid Jérôme, Pierre Hoffmeyer, Caecilia Charbonnier

Swiss Medical Weekly, 2013 , vol.  143

Accuracy assessment of Hip Clinical Exam
Article scientifique

Panayiotis Christofilopoulos, Sylvain Chagué, Schmid Jérôme, Pierre Bartolone, Pierre Hoffmeyer, Caecilia Charbonnier

Swiss Medical Weekly, 2013 , vol.  143

2012

Sensitivity of hip tissues contact evaluation to the methods used for estimating the hip joint center of rotation
Article scientifique

Ehsan Arbabi, Schmid Jérôme, Ronan Boulic, Daniel Thalmann, Nadia Magnenat-Thalmann

Medical & Biological Engineering & Computing, 2012 , vol.  50, pp.  595-604

Lien vers la publication

2011

Robust statistical shape models for MRI bone segmentation in presence of small field of view
Article scientifique

Schmid Jérôme, Jinman Kim, Nadia Magnenat-Thalmann

Medical Image Analysis, 2011 , vol.  15, no  1, pp.  155-168

Lien vers la publication

Extreme leg motion analysis of professional ballet dancers via MRI segmentation of multiple leg postures
Article scientifique

Schmid Jérôme, Jinman Kim, Nadia Magnenat-Thalmann

International Journal of Computer Assisted Radiology and Surgery, 2011 , vol.  6, pp.  47-57

Lien vers la publication

Comparison of statistical models performance in case of segmentation using a small amount of training datasets
Article scientifique

François Chung, Schmid Jérôme, Nadia Magnenat-Thalmann, Hervé Delingette

The Visual Computer, 2011 , vol.  27, pp.  141-151

Lien vers la publication

A GPU framework for parallel segmentation of volumetric images using discrete deformable models
Article scientifique

Schmid Jérôme, José Iglesias Guitián, Enrico Gobbetti, Nadia Magnenat-Thalmann

The Visual Computer, 2011 , vol.  27, pp.  85-95

Lien vers la publication

2010

Coupled registration-segmentation: Application to femur analysis with intra-subject multiple levels of detail MRI data
Article scientifique

Schmid Jérôme, Jinman Kim, Nadia Magnenat-Thalmann

MICCAI 2010. Lecture Notes in Computer Science, 2010 , vol.  6362, pp.  562-569

Collaborative telemedicine for interactive multiuser segmentation of volumetric medical images
Article scientifique

Seunghyun Han, Niels Nijdam, Schmid Jérôme, Jinman Kim, Nadia Magnenat-Thalmann

The Visual Computer, 2010 , vol.  26, pp.  639-348

Lien vers la publication

A comprehensive methodology to visualize articulations for the physiological human
Article scientifique

Nadia Magnenat-Thalmann, Schmid Jérôme, Lazhari Assassi, Pascal Volino

2010 International Conference on Cyberworlds, 2010

Lien vers la publication

2009

Musculoskeletal simulation model generation from MRI data sets and motion capture data
Article scientifique

Schmid Jérôme, Anders Sandholm, François Chung, Daniel Thalmann, Hervé Delingette, Nadia Magnenat-Thalmann

Recent Advances in the 3D Physiological Human, 2009 , pp.  3-19

Lien vers la publication

Interactive segmentation of volumetric medical images for collaborative telemedicine
Article scientifique

Schmid Jérôme, Niels Nijdam, Seunghyun Han, Jinman Kim, Nadia Magnenat-Thalmann

3DPH: Modelling the Physiological Human - Lecture Notes in Computer Science, 2009 , vol.  5903, pp.  13-24

Lien vers la publication

From MRI to anatomical simulation of the hip joint
Article scientifique

Lazhari Assassi, Caecilia Charbonnier, Schmid Jérôme, Pascal Volino, Nadia Magnenat-Thalmann

Computer Animation and Virtual Worlds, 2009 , vol.  20, no  1, pp.  53-66

Lien vers la publication

2008

Multimedia application to the simulation of human musculoskeletal system: A visual lower limb model from multimodal captured data
Article scientifique

Nadia Magnenat-Thalmann, Caecilia Charbonnier, Schmid Jérôme

2008 IEEE 10th Workshop on Multimedia Signal Processing, 2008

Lien vers la publication

MRI bone segmentation using deformable models and shape priors
Article scientifique

Schmid Jérôme, Nadia Magnenat-Thalmann

MICCAI 2008 - Lecture Notes in Computer Science, 2008 , vol.  5241, pp.  119-126

Lien vers la publication

2007

Multi-scale adaptive mask 3D rigid registration of ultrasound and CT images
Article scientifique

Zhijun Zhang, Schmid Jérôme, Man Kwan Soo, Yan Bailly, Chung Kwon Yeung

BMVC 2007 - Proceedings of the British Machine Vision Conference 2007, 2007

Lien vers la publication

Advanced da Vinci surgical system simulator for surgeon training and operation planning
Article scientifique

L.W. Sun, Frédéric Van Meer, Schmid Jérôme, Yan Bailly, A.A. Thakre, C.K. Yeung

International Journal of Medical Robotics and Computer Assisted Surgery, 2007 , vol.  3, pp.  245-251

Lien vers la publication

2024

Barriers and facilitators of implementation of AI
Conférence

Mataj-Lokaj Belinda, Pugliese Marie-Therese, K. Kinkel, C. Lovis, Schmid Jérôme

European Society of Breast Imaging (EUSOBI), 03.10.2024 - 04.10.2024, Lisbon, Portugal

Efficient clinical information extraction from breast radiology reports in French
Conférence

J. Zaghir, Mataj-Lokaj Belinda, K. Kinkel, A. Djema, H. Turbé, M. Bjelogrlic, Durand De Gevigney Valentin, Schmid Jérôme, C. Lovis, J.-P. Goldman

Medical Informatics Europe, 25.08.2024 - 29.08.2024, Athens, Greece

Enhancing radiological anatomy learning with photorealistic volumetric rendering: Insights from radiographer student education
Conférence

Schmid Jérôme, D. Domingues Rua, J. Perez Vasquez, D. Rodrigues, F.-J. Taibo Pena, J. A. Iglesias-Guitian

Swiss Congress of Radiology, 20.06.2024 - 22.06.2024, Genève, Suisse

Failure of cerebral venous sinus opacification in post-mortem CT-angiography: An artifact linked to circumstances of death
Conférence

V. Corno, M. Angulo Santander, C. Egger, Schmid Jérôme

Swiss Congress of Radiology, 20.06.2024 - 22.06.2024, Genève, Suisse

Comparative analysis of breast lesion classification with deep learning in dynamic MRI using different types of regions of interest
Conférence

Mataj-Lokaj Belinda, Durand De Gevigney Valentin, A. Djema, K. Kinkel, C. Lovis, Schmid Jérôme

Swiss Congress of Radiology, 20.06.2024 - 22.06.2024, Genève, Suisse

Enhancing breast lesion classification by integrating lesion characteristics and clinical data information with ultrafast MRI
Conférence

Mataj-Lokaj Belinda, Durand De Gevigney Valentin, J.-P. Goldman, K. Kinkel, C. Lovis, J. Zaghir, A. Djema, Schmid Jérôme

European Congress of Radiology, 28.02.2024 - 03.03.2024, Vienna, Austria

2023

Combining ultrafast MRI sequence with artificial intelligence (AI) for breast cancer detection
Conférence

Mataj-Lokaj Belinda, A. Djema, K. Kinkel, C. Lovis, Schmid Jérôme

European Congress of Radiology, 01.03.2023 - 05.03.2023, Vienna, Austria

2022

La profession TRM face à l'IA : enquête ethnographique
Conférence

Al-Musibli Azal, Schmid Jérôme

Swiss Congress of Health Professions, 01.09.2022 - 02.09.2022, Neuchâtel, Suisse

Investigating data fusion and training strategies of artificial intelligence for the diagnosis of Parkinson’s disease with Dopamine SPECT imaging
Conférence

Schmid Jérôme, A. Arrigo, G. Favre-Gillioz, N. Nicastro, V. Garibotto

European Congress of Radiology, 13.07.2022 - 17.07.2022, Vienna, Austria

Lien vers la conférence

AIRx: Augmenting the teaching of radiography using artificial intelligence
Conférence

Chênes Christophe, M. Butt, S. Mahamat, M. Muanga Bonga, S. Fazeli, Schmid Jérôme

European Congress of Radiology, 13.07.2022 - 17.07.2022, Vienna, Austria

Lien vers la conférence

Diagnosis of drowning using post-mortem computed tomography based on surface ratio and density of fluid accumulation in frontal, maxillary and sphenoid sinuses
Conférence

L. Fernandes Mendes, L. Pedreira Lago, Schmid Jérôme, C. Egger

Meeting of the Swiss Society of Forensic Medicine, 24.06.2022 - 25.06.2022, Bern, Suisse

Can artificial intelligence compete with radiographers in characterizing radiographs of the upper limb?
Conférence

Chênes Christophe, D. Locarnini, D. Perréard, B. Sierra Paulino, Schmid Jérôme

Swiss Congress of Radiology, 23.06.2022 - 25.06.2022, Fribourg, Suisse

2021

Revisiting contour-driven and knowledge-based deformable models :
Conférence ArODES
application to 2D-3D proximal femur reconstruction from X-ray images

Jerome Schmid, Christophe Chênes

Proceedings of medical image computing and computer assisted intervention – MICCAI 2021

Lien vers la conférence

Résumé:

In many clinical applications, 3D reconstruction of patient-specific structures is of major interest. Despite great effort put in 2D-3D reconstruction, gold standard bone reconstruction obtained by segmentation on CT images is still mostly used – at the expense of exposing patients to significant ionizing radiation and increased health costs. State-of-the-art 2D-3D reconstruction methods are based on non-rigid registration of digitally reconstructed radiographs (DRR) – aiming at full automation – but with varying accuracy often exceeding clinical requirements. Conversely, contour-based approaches can lead to accurate results but strongly depend on the quality of extracted contours and have been left aside in recent years. In this study, we revisit a patient-specific 2D-3D reconstruction method for the proximal femur based on contours, image cues, and knowledge-based deformable models. 3D statistical shape models were built using 199 CT scans from THA patients that were used to generate pairs of high fidelity DRRs. Convolutional neural networks were trained using the DRRs to investigate automatic contouring. Experiments were conducted on the DRRs, and calibrated radiographs of a pelvis phantom and volunteers – with an analysis of the quality of contouring and its automatization. Using manual contours and DRR, the best reconstruction error was 1.02 mm. With state-of-the-art results for 2D-3D reconstruction of the proximal femur, we highlighted the relevance and challenges of using contour-driven reconstruction to yield patient-specific models.

2020

Evaluation of an ultra-fast 4D sequence for detection of breast lesions in MRI
Conférence

Mataj-Lokaj Belinda, K. Kinkel, J.-N. Hyacinthe, Schmid Jérôme, Gaignot Céline

International Society for Magnetic Resonance in Medicine (ISMRM), 18.04.2020 - 23.04.2020, Sydney, Australia

2019

Deep learning in osteoarticular imaging : the role of the radiographer
Conférence

Al-Musibli Azal, X. Montet, O. Cagdas, Schmid Jérôme

Swiss Congress of Radiology, 13.06.2019 - 15.06.2019, St-Gallen, Suisse

2018

Towards a deformable multi-surface approach to ligamentous spine models for predictive simulation-based scoliosis surgery planning
Conférence ArODES

Michel A. Audette, Jerome Schmid, Craig Goodmurphy, Michael Polanco, Sebastian Bawab, Austin Tapp, H. Sheldon St-Clair

Computational methods and clinical applications for spine imaging

Lien vers la conférence

Résumé:

Scoliosis correction surgery is typically a highly invasive procedure that involves either an anterior or posterior release, which respectively entail the resection of ligaments and bone facets from the front or back of the spine, in order to make it sufficiently compliant to enable the correction of the deformity. In light of progress in other areas of surgery in minimally invasive therapies, orthopedic surgeons have begun envisioning computer simulation-assisted planning that could answer unprecedented what-if questions. This paper presents preliminary steps taken towards simulation-based surgery planning that will provide answers as to how much anterior or posterior release is truly necessary, provided we also establish the amplitude of surgical forces involved in corrective surgery. This question motivates us to pursue a medical image-based anatomical modeling pipeline that can support personalized finite elements simulation, based on models of the spine that not only feature vertebrae and inter-vertebral discs (IVDs), but also descriptive ligament models. This paper suggests a way of proceeding, based on the application of deformable multi-surface Simplex model applied to a CAD-based representation of the spine that makes explicit all spinal ligaments, along with vertebrae and IVDs. It presents a preliminary model-based segmentation study whereby Simplex meshes of CAD vertebrae are registered to the subject’s corresponding vertebrae in CT data, which then drives ligament and IVD model registration by aggregation of neighboring vertebral transformations. This framework also anticipates foreseen improvements in MR imaging that could achieve better contrasts in ligamentous tissues in the future.

2014

Segmentation of X-ray Images by 3D-2D Registration based on Multibody Physics
Conférence

Schmid Jérôme, Chênes Christophe

Asian Conference on Computer Vision (ACCV), 01.11.2014 - 05.11.2014, Singapore

Computer-assisted Total Hip Arthroplasty: from Pre-operative Planning to Post-operative Assessment
Conférence

Schmid Jérôme, Chênes Christophe, S. Chagué, P. Hoffmeyer, P. Christofilopoulos, C. Charbonnier

Annual Meeting of the International Society for Computer Assisted Orthopaedic Surgery (CAOS), 18.06.2014 - 21.06.2014, Milan, Italy

MyHip: Personalized Planning and Surgical Guidance in Total Hip Arthroplasty
Conférence

Schmid Jérôme, Chênes Christophe, S. Chagué, P. Christofilopoulos, C. Charbonnier, M. Bernardoni

Swiss Congress for Health Professions (SCHP), 11.03.2014 - 12.03.2014, Bern, Suisse

2013

Using Motion Capture and MRI to Accurately Determine the Hip Range of Motion in Everyday Life
Conférence

P. Christofilopoulos, S. Chagué, Schmid Jérôme, P. Bartolone, P. Hoffmeyer, C. Charbonnier

Annual Congress of the International Society of Technology in Arthroplasty (ISTA), 16.10.2013 - 19.10.2013, Miami, USA

Réalisations

Médias et communication
Nous contacter
Suivez la HES-SO
linkedin instagram facebook twitter youtube rss
univ-unita.eu www.eua.be swissuniversities.ch
Mentions légales
© 2021 - HES-SO.

HES-SO Rectorat