Workshop Program

KEYNOTE - Jean Feydy

Title: Geometric Insights on Matching Problems
Abstract: Image registration, generative modelling and cohort balancing are all matching problems: they seek an assignment that best aligns two distributions of samples under a suitable, domain-specific regularization prior. In this talk, I will summarize insights from the past decade of theoretical research in the field. In particular, I will show how objectives penalizing distortions (leading to the “Gromov-Wasserstein” or “Quadratic” Assignment Problem) can always be understood as affine registration methods in a suitable feature space. This perspective unifies several lines of work and paves the way for robust optimization algorithms that can be used either as standalone tools or as flexible layers in machine learning pipelines.

KEYNOTE - Archana Venkataraman

Title: Lightweight and Interpretable AI as a New Window into Brain Dysfunction
Abstract: Deep learning has disrupted nearly every major field, from computer vision to genomics, fueled by an explosion of data: millions of labeled images, thousands of annotated ICU admissions, hundreds of hours of transcribed speech. Clinical neuroscience is a notable holdout, a field of unavoidably small datasets, massive patient variability, and complex, largely unknown phenomena. My lab tackles these challenges from foundational neuroscientific questions to translational applications of neuroimaging and exploratory probing of neural circuitry, with a key strategy of using domain knowledge to build models that are both lightweight and interpretable. This talk highlights two ongoing lines of work in epilepsy. First, seizure detection from scalp EEG, where a simple transformer combines spatial and temporal information in continuous recordings to pinpoint both the time of onset and the involved scalp areas across a large clinical dataset. Second, as a follow-up, we ask what attributes of the EEG signal drive the model’s prediction: using contrastive training we align the model’s EEG encodings with textual concept embeddings derived from clinical notes, then apply attention-weighted pooling to detect patient-specific seizure and baseline etiologies.

PROGRAMME

MLCN 2026 @ MICCAI 2026, Strasbourg, France. Thursday, 1 October 2026 (all times local).

Time	Program	Topic	Presenter	Institution
1:30 PM	Opening
Session 1: ML
1:35 PM	Keynote #1	Geometric Insights on Matching Problems	Jean Feydy	Inria Paris, France
2:15 PM	Lightning Talks	Text-Guided Multimodal Multitask Learning for Brain Tumor Segmentation	Mumu Aktar	Toronto Metropolitan University, Canada
		Rapid Whole-Brain Parcellation	Joshua R. Astley	King's College London, UK
		Selection of Informative Variables in Y-Aware Framework: Insights from Genetic and Behavioural Data	Antoine Didier	NeuroSpin, France
		Representation Learning for 3D Brain Imaging: A Benchmark	Pierre Falconnier	INSA Lyon, France
		Few-Shot Cross-Site Domain Generalization for Multi-Site Autism Brain Network Classification	Fatima Ez-Zahraa Bazay	Mohammed V University, Morocco
		Do CNNs Learn Clinically Meaningful Imaging Representations? A Portable Multi-level Representation Audit for Brain Tumor MRI Classification	Bo-Wei Lai	National Yang Ming Chiao Tung University, Taiwan
		MRI-Based Brain Age Estimation with Supervised Contrastive Learning of Continuous Representation	Simon Joseph Clément Crête	Concordia University, Canada
		EEG-LoGNet: Bridging Local Features and Global Contexts for EEG-Based Motor Imagery Classification	Jinsong Guo	Southern University of Science and Technology, China
		Heterogeneous Reservoir Dynamics Reveal Disease-Specific Temporal Fingerprints	Mayssa Soussia	ENISo, Tunisia
2:25 PM	Oral 1	MEROS: Multi-view Expert Routing in Ordinal Space for MDD Spectrum Assessment	Yilin Leng	University of Science and Technology of China
2:35 PM	Oral 2	SARAR: Shortcut-Aware 3D Brain MRI Question Answering via Retrieval-Augmented Reranking	Mohammad H. Abbasi	Stanford University, USA
2:45 PM	Oral 3	Spatial Feature-wise Linear Modulation (SpFiLM) for Contrast Agent-Aware Brain Parcellation	Pushpendra Singh	King's College London, UK
2:55 PM	Oral 4	TIIC: Tabular Integration of Imaging and Clinical Data for Interpretable Multimodal Inference for Alzheimer’s Disease	Chaima Hammami	University of Strasbourg, France
Poster Session
3:05 PM	Poster Session
3:30 PM	Coffee Break
Session 2: CN
4:00 PM	Keynote #2	Lightweight and Interpretable AI as a New Window into Brain Dysfunction	Archana Venkataraman	Boston University, USA
4:40 PM	Lightning Talks	Probability-Invariant Random Walk Learning on Gyral Folding-Based Cortical Similarity Networks for Alzheimer's and Lewy Body Dementia Diagnosis	Minheng Chen	University of Texas at Arlington, USA
		SFINX: Structure-informed Functional-MRI Integration via xLSTM for Autism Diagnosis	Peiyu Duan	Yale University, USA
		Automated Segmentation and Height Measurement of Pituitary Gland	Vahe Petrosyan	American University of Armenia
		Toward Personalized Dyslexia Classification via Dynamic Functional Connectivity and Explainable AI	Blessy Thomas	University of Qatar
		The Diagnosis a Reporter Leaves Unspoken: Surfacing Frozen Tumor Features for Brain-Tumor MRI Reporting	Khawaja Murad ul Hassan	National University of Science and Technology, Pakistan
		From Blood to Brain: Uncertainty-Aware Adaptive Fusion for Alzheimer’s Staging and Progression Under Incomplete Multimodal Profiles	Shahzad Ali	University of Bologna, Italy
		Network Alterations Precede Atrophy in Alzheimer’s Disease Progression Subtypes	Bailey Berman	Erasmus MC, The Netherlands
		Graph-Theoretical Task-Based Frontal Brain Network of Visual Working-Memory Load: An fNIRS Study	Arman Nik Khah	University of Texas at Dallas, USA
		Implicit Neural Representations for Modeling the Accumulation of Tau Protein in the Brain	Nicolas Honnorat	University of Texas Health Science Center at San Antonio, USA
4:50 PM	Oral 5	A Unified Brain MRI Reporting Framework Built on CoT-Guided VLM and Expert Models	Yuxiao Liu	ShanghaiTech University, China
5:00 PM	Oral 6	Paramagnetic Rim Lesion Instance Segmentation in Multiple Sclerosis Using Conditional Convolutions	Amirhossein Rasoulian	NeuroRx, Canada
5:10 PM	Oral 7	What Do Persistent Misclassifications Tell Us About Alzheimer’s Disease Detection using Structural MRI?	Didem Stark	University of Tübingen, Germany
5:20 PM	Oral 8	Cross-Modality Structural Guidance in 3D Latent Diffusion for Robust FLAIR Super-Resolution	Andrew Marshall (on behalf of Haoyu Lan)	Yale University, USA
5:30 PM	Closing + Awards
6:00 PM	End