Publications

GeoLS: an Intensity-based, Geodesic Soft Labeling for Image Segmentation
Sukesh Adiga Vasudeva
Jose Dolz
Hervé Lombaert
Soft-label assignments have emerged as prominent strategies in training dense prediction problems, such as image segmentation. These approac… (see more)hes mitigate the limitations of hard labels, such as inter-class relationships in the data and spatial relationships between a given pixel and its neighbors. Nevertheless, most existing methods rely only on ground-truth masks and ignore the underlying image context associated with each label. For instance, image intensities convey information that could potentially clear ambiguities in the annotation. This paper, therefore, proposes a Geodesic Label Smoothing (GeoLS) approach that incorporates image intensity information within the soft labeling process. Specifically, we leverage the geodesic distance transform to capture the intensity variations between pixels. The generated maps geodesically modify the hard labels to obtain new intensity-based soft labels. The resulting geodesic soft labels better model spatial and class-wise relationships as they capture the variations of image gradients across classes and anatomy. The benefits of our intensity-based geodesic soft labels are assessed on three diverse sets of publicly accessible segmentation datasets. Our experimental results show that the proposed method consistently improves the segmentation accuracy compared to state-of-the-art soft-labeling techniques in terms of the Dice similarity and Hausdorff distance.
2025-04-17
Machine Learning for Biomedical Imaging (published)
doi.org
Multimodal population study reveals the neurobiological underpinnings of chronotype
Le Zhou
Karin Saltoun
Julie Carrier
Kai-Florian Storch
Robin I. M. Dunbar
Danilo Bzdok
2025-04-16
Nature Human Behaviour (unknown)
doi.org
Alcohol related hepatitis in intensive care units: clinical and biological spectrum and mortality risk factors: a multicenter retrospective study
Maxime Gasperment
Léa Duhaut
Nicolas Terzi
Côme Gerard
Luc Haudebourg
Alexandre Demoule
Mialy Randrianarisoa
Vincent Castelain
Sacha Sarfati
Fabienne Tamion
Charlene Le Moal
Christophe Guitton
Gabriel Preda
Arnaud Galbois
Thibault Vieille
Gaël Piton
Marika Rudler
Guillaume Dumas
Hafid AIT-OUFELLA
Background Alcohol related hepatitis is responsible for high morbidity and mortality, but little is known about the management of patients w… (see more)ith hepatitis specifically in intensive care units (ICU). Methods Retrospective study including patients with alcohol related hepatitis hospitalized in 9 French ICUs (2006–2017). Alcohol related hepatitis was defined histologically or by an association of clinical and biological criteria according to current guidelines. Results 187 patients (median age: 53 [43–60]; male: 69%) were included. A liver biopsy was performed in 51% of cases. Patients were admitted for impaired consciousness (71%), sepsis (64%), shock (44%), respiratory failure (37%). At admission, median SOFA and MELD scores were 10 [7–13] and 31 [26–40] respectively. 63% of patients received invasive mechanical ventilation, 62% vasopressors, and 36% renal replacement therapy. 66% of patients received corticosteroids, and liver transplantation was performed in 16 patients (8.5%). ICU and in-hospital mortality were 37% and 53% respectively. By multivariate analysis, ICU mortality was associated with SOFA score (without total bilirubin) (SHR 1.08 [1.02–1.14] per one-point increase), arterial lactate (SHR 1.08 [1.03–1.13] per 1 mmol/L) and MELD score (SHR 1.09 [1.04–1.14] per 1 point), while employment was associated with increased survival (HR 0.49 [0.28–0.86]). After propensity score weighting, the use of corticosteroids did not affect ICU mortality in the overall population but had a beneficial effect in the subgroup of patients with histological proof. Patient prognosis was also better in responders assessed by Lille score at day 7 (OR 6.67 [2.44–20.15], p  0.001). Conclusion Alcohol related hepatitis is a severe condition leading to high mortality in ICU patients. Severity of organ failure
2025-04-14
Annals of Intensive Care (published)
doi.org
Assessing Critical Thinking Skills in Data Literacy: A Digital Performance Assessment
Ying Cui
Xiaoxiao Liu
Fu Chen
Alina Lutsyk
Jaqueline P. Leighton
Maria Cutumisu
2025-04-14
Zeitschrift für Pädagogik (published)
doi.org
Brain Diffusion Transformer for Personalized Neuroscience and Psychiatry
Rongquan Zhai
Yechen Hu
Liping Zheng
Shitong Xiang
Chao Xie
Lei Peng
Tobias Banaschewski
Gareth J. Barker
Arun L.W. Bokde
Rüdiger Brühl
Sylvane Desrivières
Herta Flor
Hugh Garavan
Penny Gowland
Antoine Grigis
Andreas Heinz
Herve Lemaitre
Jean-Luc Martinot
Marie-Laure Paillère Martinot
Eric Artiges … (see 26 more)
Frauke Nees
Dimitri Papadopoulos Orfanos
Luise Poustka
Michael N. Smolka
Sarah Hohmann
Nathalie Holz
Nilakshi Vaidya
Robert Whelan
Zuo Zhang
Lauren Robinson
Jeanne Winterer
Sinead King
Yuning Zhang
Hedi Kebir
Ulrike Schmidt
Julia Sinclair
Argyris Stringaris
Gunter Schumann
Danilo Bzdok
Henrik Walter
Edmund T. Rolls
Barbara Sahakian
Trevor W. Robbins
Jianfeng Feng
Weikang Gong
Tianye Jia
Task-fMRI analyses typically focus on localized activation contrasts between stimuli, neglecting the brain’s dynamic hierarchy. We introdu… (see more)ce Brain Diffusion Transformer (Brain-DiT), a deep generative model capturing recurrent processing underlying individualized neurocognitive state transitions via functional networks. Without prior assumptions, Brain-DiT identifies canonical cognitive regions in the brain and reveals replicable subgroups with distinct neural circuits in large cohorts, offering critical clinical insights overlooked by traditional methods: individuals exhibiting negative emotion bias, linked to language-related regions, had a 12-fold higher likelihood of major depression, and those with maladaptive inhibition strategies, associated with overactive medial frontal regions, showed a 9-fold increased risk of alcohol abuse. By bridging cognitive theory and psychiatric applications, Brain-DiT provides a unified analytical paradigm, paving the way for operational personalized medicine in psychiatry.
2025-04-13
bioRxiv (preprint)
doi.org
IL1RAP is an immunotherapeutic target for normal karyotype triple-mutated acute myeloid leukemia
Arnaud Metois
Marie-Eve Bordeleau
Louis Theret
Azadeh Hajmirza
Ossama Moujaber
Jean-François Spinella
Jalila Chagraoui
Nadine Mayotte
Isabel Boivin
Éric Audemard
Léo Aubert
Véronique Lisi
Banafsheh Khakipoor
Azer Farah
Éric Bonneil
Alma Robert
Julie Lippens
Anna Moraitis
Francois Béliveau
Albert Feghaly … (see 10 more)
Geneviève Boucher
Richard Marcotte
Patrick Gendron
Pierre Thibault
Sébastien Lemieux
Guillaume Richard-Carpentier
Vincent-Philippe Lavallee
Josée Hébert
Philippe P. Roux
Guy Sauvageau
Surface antigens of potential clinical significance remain under-characterized in AML. The European Leukemia Network classifies normal karyo… (see more)type AML (NK-AML) mutated for NPM1 (NPM1c) as a distinct entity associated with favorable outcomes if not associated with FLT3-ITD mutation. A subset of NPM1c NK-AML shows additional mutations in 2 genes: FLT3 (FLT3-ITD) and DNMT3 A. These leukemias, also referred to as NK triple mutated AML (NKt-AML), are particularly difficult to eradicate with current treatment options. Therefore, novel therapies are necessary that use proteins specifically expressed at the surface. In order to identify surface antigens for immunotherapy in NKt-AML, an extensive multi-omic analysis was conducted on primary AML samples. Surface proteome enrichment was performed on 100 primary AML samples, twelve of which were NKt-AML. Transcriptome analysis was carried out on the 691 primary AML samples, and single-cell RNA sequencing was conducted on 23 primary AML samples. Herein, using multi-omics data from the Leucegene collection, we identify IL1RAP as a promising antigen for this AML subgroup. We demonstrate that IL1RAP is expressed at the surface of primitive AML cells reminiscent of leukemic stem cells in NKt-AML primary human AML specimens, while showing relatively low expression levels in normal bone marrow HSCs. Furthermore, results indicate that elevated IL1RAP expression associates with poor overall and relapse-free survival in the Leucegene cohort of AML patients and predicts nonresponse to hematopoietic stem cell transplantation. Finally, we show that IL1RAP protein is internalized following exposure to specific antibodies, suggesting that IL1RAP represents an interesting target for antibody–drug conjugate development in NKt-AML. IL1RAP exhibits preferential expression within NKt-AML, correlating with diminished overall survival rates and diminished responsiveness to hematopoietic stem cell transplantation. Moreover, internalization of IL1RAP presents a promising avenue for immunotherapeutic intervention. The online version contains supplementary material available at 10.1186/s40364-025-00769-z.
2025-04-13
Biomarker Research (published)
doi.org
Improving Quality Control of MRI Images Using Synthetic Motion Data
C Bricout
K Cho
M Harms
O Pasternak
C Bearden
PD McGorry
RS Kahn
JM Kane
B Nelson
SW Woods
ME Shenton
S Bouix
S Ebrahimi Kahou
MRI quality control (QC) is challenging due to unbalanced and limited datasets, as well as subjective scoring, which hin-der the development… (see more) of reliable automated QC systems. To address these issues, we introduce an approach that pretrains a model on synthetically generated motion artifacts before applying transfer learning for QC classification. This method not only improves the accuracy in identifying poor-quality scans but also reduces training time and resource requirements compared to training from scratch. By leveraging syn-thetic data, we provide a more robust and resource-efficient solution for QC automation in MRI, paving the way for broader adoption in diverse research settings.
2025-04-13
2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI) (published)
doi.org
arxiv.org
Logging Requirement for Continuous Auditing of Responsible Machine Learning-based Applications
Patrick Loic Foalem
Leuson Da Silva
Foutse Khomh
Heng Li
Ettore Merlo
Machine learning (ML) is increasingly applied across industries to automate decision-making, but concerns about ethical and legal compliance… (see more) remain due to limited transparency, fairness, and accountability. Monitoring through logging a long-standing practice in traditional software offers a potential means for auditing ML applications, as logs provide traceable records of system behavior useful for debugging, performance analysis, and continuous auditing. systematically auditing models for compliance or accountability. The findings underscore the need for enhanced logging practices and tooling that systematically integrate responsible AI metrics. Such practices would support the development of auditable, transparent, and ethically responsible ML systems, aligning with growing regulatory requirements and societal expectations. By highlighting specific deficiencies and opportunities, this work provides actionable guidance for both practitioners and tool developers seeking to strengthen the accountability and trustworthiness of ML applications.
2025-04-13
Empirical Software Engineering (published)
doi.org
arxiv.org
Open Problems in Technical AI Governance
Anka Reuel
Benjamin Bucknall
Stephen Casper
Timothy Fist
Lisa Soder
Onni Aarne
Lewis Hammond
Lujain Ibrahim
Alan Chan
Peter Wills
Markus Anderljung
Ben Garfinkel
Lennart Heim
Andrew Trask
Gabriel Mukobi
Rylan Schaeffer
Mauricio Baker
Sara Hooker
Irene Solaiman
Sasha Luccioni … (see 14 more)
Alexandra Luccioni
Nitarshan Rajkumar
Nicolas Moës
Jeffrey Ladish
David Bau
Paul Bricman
Neel Guha
Jessica Newman
Yoshua Bengio
Tobin South
Alex Pentland
Sanmi Koyejo
Mykel Kochenderfer
Robert Trager
AI progress is creating a growing range of risks and opportunities, but it is often unclear how they should be navigated. In many cases, the… (see more) barriers and uncertainties faced are at least partly technical. Technical AI governance, referring to technical analysis and tools for supporting the effective governance of AI, seeks to address such challenges. It can help to (a) identify areas where intervention is needed, (b) identify and assess the efficacy of potential governance actions, and (c) enhance governance options by designing mechanisms for enforcement, incentivization, or compliance. In this paper, we explain what technical AI governance is, why it is important, and present a taxonomy and incomplete catalog of its open problems. This paper is intended as a resource for technical researchers or research funders looking to contribute to AI governance.
2025-04-13
Transactions on Machine Learning Research (accepted)
doi.org
openreview.net
Predicting College Enrollment for Low-Socioeconomic-Status Students Using Machine Learning Approaches
Surina He
Mehrdad Yousefpoori-Naeim
Ying Cui
Maria Cutumisu
2025-04-11
Big Data and Cognitive Computing (published)
doi.org
BindGPT: A Scalable Framework for 3D Molecular Design via Language Modeling and Reinforcement Learning
Artem Zholus
Maksim Kuznetsov
Roman Schutski
Shayakhmetov Rim
Daniil Polykovskiy
A. Chandar
Alex Zhavoronkov
Generating novel active molecules for a given protein is an extremely challenging task for generative models that requires an understanding … (see more)of the complex physical interactions between the molecule and its environment. In this paper, we present a novel generative model, BindGPT which uses a conceptually simple but powerful approach to create 3D molecules within the protein's binding site. Our model produces molecular graphs and conformations jointly, eliminating the need for an extra graph reconstruction step. We pretrain BindGPT on a large-scale dataset and fine-tune it with reinforcement learning using scores from external simulation software. We demonstrate how a single pretrained language model can serve at the same time as a 3D molecular generative model, conformer generator conditioned on the molecular graph, and a pocket-conditioned 3D molecule generator. Notably, the model does not make any representational equivariance assumptions about the domain of generation. We show how such simple conceptual approach combined with pretraining and scaling can perform on par or better than the current best specialized diffusion models, language models, and graph neural networks while being two orders of magnitude cheaper to sample.
2025-04-10
Proceedings of the AAAI Conference on Artificial Intelligence (published)
doi.org
arxiv.org
Designing Ambiguity Sets for Distributionally Robust Optimization Using Structural Causal Optimal Transport
Ahmad-reza Ehyaei
Golnoosh Farnadi
Samira Samadi
Distributionally robust optimization tackles out-of-sample issues like overfitting and distribution shifts by adopting an adversarial approa… (see more)ch over a range of possible data distributions, known as the ambiguity set. To balance conservatism and accuracy, these sets must include realistic probability distributions by leveraging information from the nominal distribution. Assuming that nominal distributions arise from a structural causal model with a directed acyclic graph
2025-04-10
Proceedings of the AAAI Conference on Artificial Intelligence (published)
doi.org
arxiv.org