Olivier Colliot
@oliviercolliot
Researcher @AramisLabParis @CNRS @Inria @Inserm @Sorbonne_Univ_ @InstitutCerveau @InstitutPrairie - Machine learning, medical imaging, brain disorders
"Machine Learning for Brain Disorders" is now published! 1058 pages fully open access link.springer.com/book/10.1007/9…
Excited to share nnInteractive: The nnU-Net moment for 3D Interactive Segmentation! Supports points, scribbles, boxes, & a new lasso prompt for full 3D masks from 2D inputs. Check it out: arxiv.org/abs/2503.08373 github.com/MIC-DKFZ/napar… mitk.org/wiki/MITK-nnIn…
Across MICCAI 2023 segmentation papers: Median improvement over SOTA: 1 percent point of Dice Median (imputed) Confidence Interval width: 3 percent points Should we be worried? 😱 To learn more, see poster T-AM-065 by @evangelia_chris @annika_reinke1 at 10.30 #MICCAI2024
🔥#MICCAI2024 Teaser 1🔥: Can we trust our medical imaging AI results? We analyzed all 221 MICCAI 2023 segmentation papers to answer this: arxiv.org/pdf/2409.17763 Check out the poster presented by @evangelia_chris @annika_reinke1 on Oct 8th 10:30. Great collaboration…
Back to #MICCAI2024 after some years is really exciting, great job of the team with posters and presentations. And it just started! @MiccaiStudents @MICCAI_Society @CIBM_ch @HaslerStiftung @fns_ch @HamzaKebiri @RizhongLin @FBM_UNIL
Welcome to #MICCAI2024! Today marks the beginning of an exciting series of workshops, challenges, and tutorials. We're happy you're here! @MiccaiStudents @WomenInMICCAI @RMiccai @ja_schnabel @KarimLekadir
🔥#MICCAI2024 Teaser 1🔥: Can we trust our medical imaging AI results? We analyzed all 221 MICCAI 2023 segmentation papers to answer this: arxiv.org/pdf/2409.17763 Check out the poster presented by @evangelia_chris @annika_reinke1 on Oct 8th 10:30. Great collaboration…
… between @helmholtz_image the @MICCAI_Society SIG on challenges @bias_sig the @ProjectMONAI benchmarking group and involving a great team of researchers (see pic). Shared senior authors @oliviercolliot @GaelVaroquaux @lena_maierhein @ellislifehd
🙌We are beyond excited to have the great @lena_maierhein as our keynote speaker at SASHIMI 2024! 🙌 #MICCAI2024 Keynote title: "Medical Image Synthesis for Data-Centric AI: Lessons Learnt" ✨ Come join at 13:30! ✨ More info: 2024.sashimi-workshop.org/keynote/
🚨 New publication alert: 📢 “Automated MRI Quality Assessment of Brain T1-weighted MRI in Clinical Data Warehouses: A Transfer Learning Approach Relying on Artefact Simulation.” 🖊️ @loizillon_s, S. Bottani, S. Mabille, Y. Jacob, ... , @oliviercolliot, @NinonBurgos et al. ⬇️
Découvrez le projet APPRIMAGE qui vise à valider à grande échelle une méthode d’apprentissage automatique pour l’aide au diagnostic de maladies neurologiques à partir de données d’IRM cérébrale bernoulli-lab.fr/project/apprim… linkedin.com/posts/bernoull…
Happy to share this cross-diagnostic review paper on Dimensional Neuroimaging Endophenotypes: Neurobiological Representations of Disease Heterogeneity Through Machine Learning sciencedirect.com/science/articl… Led by @JunhaoWen
🚨 New publication alert: 📢 “Evaluation of pseudo-healthy image reconstruction for anomaly detection with deep generative models: Application to brain FDG PET.” 🖊️ @HassanalyRavi, C. Brianceau, @maelysolal, @oliviercolliot, @NinonBurgos. 🎯 Authors propose an evaluation…
In 2021, we published nnU-Net addressing an ongoing bias towards novel architectures in medical image segmentation. Now, 3 years later, we observe that this innovation bias prevails, and - despite recent claims - scaled U-Net variants remain SotA, at least on current benchmarks.
Concerning results: We've evaluated Transformer-, Mamba-, and CNN-based architectures, revealing widespread validation flaws in #MedicalImageSegmentation. Many claims of superiority do not withstand strict testing. Time to push for rigor in the field :) arxiv.org/pdf/2404.09556…
Merci à tous les collègues grenoblois pour l'organisation de cette magnifique nouvelle édition d'IABM ! En particulier @benlemasson @MichelDojat @MIAI_UGA