Portrait of Martin Vallières

Martin Vallières

Associate Academic Member
Associate Professor, McGill University
Research Topics
AI and Healthcare
Medical Machine Learning

Biography

Martin Vallières is a researcher developing AI methods for Precision Medicine. His research expertise is at the intersection of AI and clinical sciences. His distinct research background contributed to the development of this “hybrid” type of expertise, one of key importance to accelerate the adoption of AI methods in the clinical environment.

Martin Vallières studied Physics Engineering at the Bachelor level. From 2010 to 2017, he then studied Medical Physics at the MSc and PhD levels and developed multiple predictive models for different cancer types. From 2017 to 2020, he pursued different postdoctoral internships in which he developed multimodal predictive models in oncology. On April 2020, he joined the Department of Computer Science at Université de Sherbrooke as an Assistant Professor and a Canada CIFAR AI Chair.

On August 2025, Martin Vallières changed affiliation and was appointed Associate Professor at the Medical Physics Unit of the Department of Oncology of McGill University. This new appointment will allow Martin Vallières to be in closer relation with clinical research teams and health domain end-users, a key point for the success of his research program.

Current Students

PhD - McGill University
Co-supervisor :
PhD - Université de Sherbrooke
Principal supervisor :
Postdoctorate - McGill University
Co-supervisor :
Master's Research - McGill University
Co-supervisor :

Publications

Medical Image Analysis with Machine Learning (MEDiml): Strategies for Simpler Radiomics
Élodie Gillard
Moustafa Amine Bezzahi
Alexander Chum
Ben Coull-Neveu
Mariem Kallel
Oliver Lefebvre
Nicolas Longchamps
Thanh-Tri Nguyen
Ouael Nedjem Eddine Sahbi
Jérémy Dana
Radiomics extracts quantitative features describing tumor shape, intensity, and texture from medical images to build models that predict cli… (see more)nical outcomes. However, large feature sets increase the risk of overfitting in small datasets, and building an end-to-end workflow often requires programming expertise. We present MEDiml (Medical Image Analysis with Machine Learning), open-source radiomics software available as a Python package (MEDiml-pkg) and a no-code desktop application (MEDiml-app). Both run on the same computational engine, and code generation exports graphically configured workflows as executable Python code. Feature extraction follows the Image Biomarker Standardization Initiative (IBSI) standards. MEDiml is built around complexity minimization: features are grouped into levels of increasing complexity (morphological, intensity, textural,l and filter-based), a model is trained at each level, and the simplest level not significantly outperformed by more complex ones is retained. Applied to five oncological datasets (2,104 scans), MEDiml achieved AUCs of 0.65 to 0.82, with morphological or intensity features sufficient in four of the five tasks. Resources: www.mediml.app.
DIGOM: Disentangled Interaction Graphs for Outcome prediction in Multimodal healthcare
Multimodal healthcare data contain relational information that extends beyond individual patient observations. Effective clinical outcome pr… (see more)ediction therefore requires not only integrating information from multiple modalities, but also modeling the relationships between them. Standard fusion approaches do not explicitly capture these interactions, while existing graph-based approaches typically model only a subset of them. We propose DIGOM (Disentangled Interaction Graphs for Outcome prediction in Multimodal healthcare), a graph-based framework that jointly models three complementary interaction types: Intra-patient cross-modal, Cross-patient same-modal, and Cross-patient cross-modal interactions. DIGOM learns a sparse multiplex graph with disentangled views, each encoding interactions between pairs of modalities across patients. Discrete adjacency matrices with patient-adaptive connectivity are sampled via a differentiable mechanism, enabling end-to-end learning and allowing the model to select informative relationships without imposing a fixed number of neighbors. A regularization objective maintains overall graph sparsity, while message passing integrates this relational context into patient representations for outcome prediction. We evaluate DIGOM on two multimodal healthcare datasets, MIMIC-IV and NIH-TB, across five clinical prediction tasks. DIGOM achieves the highest AUROC on all five tasks while using fewer connections than existing multi-interaction approaches. Ablation studies reveal that each interaction type's contribution varies across tasks, highlighting the need to adapt relational structures per task. DIGOM also demonstrates greater robustness to missing modalities, suggesting learned relationships provide complementary information when modalities are unavailable. Post-hoc analyses further reveal task- and outcome-specific connectivity patterns in the learned graphs.
Every Edge Needs a Quote: Building a Quote-Grounded Knowledge Graph of Healthcare Software Adoption with Language-Model Agents
Juan Felipe Duran
Arthur Tobler
Hospitals regularly buy software that clinicians and patients depend on. Because these purchases are costly, hard to reverse and can affect … (see more)patient care, buyers look for evidence of which peer institutions already use a product and which systems it connects to. This evidence is public but scattered across vendor websites, press releases and third-party pages. Language-model agents can now collect it at low cost, but the correctness of the extracted claims is difficult to establish, and a wrongly listed customer can mislead buyers and misrepresent an organisation. We present a pipeline that builds a knowledge graph of customer and integration relations from the open web under one rule: every edge must be supported by a verbatim quote from its source page. Pages come from a deterministic crawl of each vendor's site and a budgeted ReAct agent that searches the rest of the web. A permissive extractor proposes relations with quotes, and a layered grounding stage decides which to keep, using a referent step that reads the whole page when a quote names only one side, deterministic checks, and a panel of five model calls voting on nine criteria. On 61 vendors, the pipeline reaches 0.94 precision and 0.71 recall against a page-level key covering the 6,428 candidate relations proposed by its extractor, whose verdicts were drafted by one model and re-checked with a second. Against the golden reference, the marketplace's own curated list, it scores lower, but 89% of the relations it keeps that the reference lacks are stated on a page. A confidence score built from the panel's votes keeps more correct relations than the fixed rule at the same precision, and reaches 0.90 precision at 0.90 recall. We also discuss how narrow, bounded model calls, string-matched evidence and voting address long-context degradation and run-to-run variance.
Reference Manual of MEDiml
Élodie Gillard
Moustafa Amine Bezzahi
Alexander Chum
Ben Coull-Neveu
Mariem Kallel
Oliver Lefebvre
Nicolas Longchamps
Thanh-Tri Nguyen
Ouael Nedjem Eddine Sahbi
Jérémy Dana
Radiomics is the high-throughput extraction of quantitative features (e.g., tumor shape, intensity distributions, and texture) from medical … (see more)images to characterize tissues and uncover disease patterns invisible to the human eye. These features can serve as candidate imaging biomarkers for cancer diagnosis, prognosis, and classification. However, extracting thousands of features makes it difficult to identify the most informative ones for a given clinical task, while implementing an end-to-end workflow can require programming expertise that limits accessibility for clinical researchers. Hence, we propose MEDiml, an open-source, standards-compliant radiomics platform connecting feature extraction, predictive modeling, and interpretation through complementary graphical and programmatic interfaces. MEDiml combines a Python package (MEDiml-pkg) with a graphical desktop application (MEDiml-app) that enables researchers to conduct experiments without writing code. Both interfaces use the same computational implementation, and code generation exports graphically configured workflows as executable Python code, facilitating reproducibility and collaboration between clinical researchers and computational scientists. A typical MEDiml workflow includes image preprocessing, radiomic feature extraction, feature set reduction, model training and evaluation, and interpretation using feature importance. Feature sets can be grouped into levels of radiomic complexity (morphological, intensity-based, textural, and filter-based) to evaluate whether more complex features provide additional predictive value. This reference manual documents the radiomics concepts, workflow, software components, architecture, and installation, compares MEDiml with existing tools, and demonstrates its complexity-driven workflow on five oncological imaging cohorts comprising 2,104 patients. Resources: www.mediml.app.
Diagnosing Low-Confidence Patient Profiles via Localized Harmful Covariate Shift Detection
Before deploying clinical prediction models, their reliability must be assessed to ensure safe use. Errors may arise from distribution shift… (see more)s between training and deployment or from heterogeneous performance where the model systematically underperforms for certain patient subgroups. Yet most post-hoc diagnostic methods evaluate reliability only at the population level, checking whether the deployment data resemble the training data and whether the model performs well on average, without accounting for heterogeneity that may cause some patients to be disproportionately affected. We propose a unified diagnostic pipeline that combines both perspectives. MED3pa first discovers interpretable low-confidence patient profiles, and Detectron is then applied within each profile to test whether its degradation is associated with a detectable harmful covariate shift. Across three clinical prediction tasks and three injected subgroup shifts in MIMIC-IV, we find that shift detectability itself is informative. An injected age shift, undetectable at the population level, was nonetheless localized to specific degraded profiles, whereas sex- and race-related shifts were detected broadly and therefore offered little guidance for identifying the most affected patients.
AN AGENTIC FRAMEWORK FOR INTERROGATABLE AND INTERPRETABLE SURVIVAL ANALYSIS
Juan Felipe Duran
Shirin Abbasinejad Enger
Modern Convolutional Design Improves Uterine MRI Segmentation, while nnU-Net Remains Most Robust Across Heterogeneous Datasets
Daniel A. Di Giovanni
Akiyo Takada
Evan McNabb
Jérémy Dana
Hajime Yokota
Takahiro Tsuboyama
Rita Zakarian
James Man Git Tsui
Caroline Reinhold
Abstract Purpose To evaluate how segmentation architecture and dataset-adaptive configuration influence uterine MRI segmentation across hete… (see more)rogeneous benign and malignant tasks. Methods U-Net, Swin-UNETR, and MedNeXt were compared with nnU-Net as a self-configuring reference across T2-weighted MRI datasets: public multiclass UMD anatomy/fibroid segmentation (n=300), institutional endometrial cancer tumor segmentation (n=206), and institutional uterine mass lesion segmentation (n=234). A relabeled external UMD-style cohort (n=12) assessed domain shift. Models used fixed partitions, fold ensembling, Dice, HD95, ASSD, volume error, and paired bootstrap comparisons with Holm correction. Results MedNeXt was the strongest manually controlled architecture. nnU-Net achieved the highest performance on all internal datasets and external testing. Macro-Dice reached 0.761, 0.746, and 0.814 for nnU-Net on UMD, endometrial cancer, and uterine mass datasets, respectively, versus 0.722, 0.726, and 0.789 for MedNeXt. The nnU-Net-MedNeXt gap was largest for multiclass UMD segmentation and smaller in binary tasks. External testing degraded all models; nnU-Net remained highest (0.542), followed by MedNeXt (0.490), U-Net (0.396), and Swin-UNETR (0.287). Conclusions Uterine MRI segmentation performance depended on task, architecture, and evaluation domain. MedNeXt supported modern convolutional design as a strong manual baseline, but nnU-Net remained the most robust overall, emphasizing the importance of dataset-adaptive configuration and external validation.
Across-Site MRI Prediction of Substantial Lymphovascular Space Invasion in Endometrial Cancer: Radiomics versus Deep Learning Features
Daniel A. Di Giovanni
Akiyo Tanaka
Takuro Horikoshi
Takahiro Tsuboyama
H Yokota
Rita Zakarian
Yuka Matsumoto
Caroline Reinhold
Purpose: To compare the cross-site generalization of radiomic features and deep learning embeddings for MRI prediction of substantial lympho… (see more)vascular space invasion (LVSI) in endometrial cancer. Materials and Methods: This retrospective two-center study included 206 women (mean age, 59.8 years) with endometrial cancer who underwent preoperative 3-T MRI from March 2016 to March 2023. Hospital A (n = 130) was used for development and Hospital B (n = 76) for strict external testing. T2-weighted, reduced field-of-view diffusion-weighted, and apparent diffusion coefficient images were manually segmented. Radiomic features and seed-pooled embeddings from 3D ResNet18, DenseNet121, and U-NEXtractor were modeled with elastic-net logistic regression or XGBoost. Out-of-fold Platt calibration and sensitivity-targeted thresholds were estimated using development data only. AUCs were summarized with 95% bootstrap confidence intervals. Results: External radiomics with elastic-net achieved an AUC of 0.609 (95% CI: 0.464, 0.740) and sensitivity of 0 of 12 (0%). DenseNet121 with elastic-net had the highest external AUC (0.685; 95% CI: 0.538, 0.822) but sensitivity of 3 of 12 (25%). U-NEXtractor with elastic-net detected 10 of 12 positive cases (83.3%) with specificity of 32 of 64 (50.0%) and balanced accuracy of 0.667. XGBoost showed higher apparent development performance but weaker external operating behavior. Conclusion: Under real-world cross-site MRI acquisition shift, DenseNet121 and U-NEXtractor embeddings showed better external generalization than handcrafted radiomic features for substantial LVSI prediction.
Radiomic prediction of substantial LVSI in endometrial cancer using reduced field of view DWI- a feasibility study
Akiyo Takada
Daniel A. Di Giovanni
Takuro Horikoshi
Takahiro Tsuboyama
Hajime Yokota
Sakurako Harada‐Kagitani
Evan McNabb
Jérémy Dana
Rita Zakarian
Haruto Sugawara
Yuka Matsumoto
Yuji Habu
Hirokazu Usui
Kaori Koga
Katsuhiro Nasu
Takashi Uno
Caroline Reinhold
Optimizing renal mass histology prediction with artificial intelligence using handcrafted radiomics
Teodora Boblea Podasca
Marc-Antoine Blais
Marie-Lou Gadoury-Campbell
Stéphanie Boulet
Patrick O. Richard
A Multimodal and Multi-centric Head and Neck Cancer Dataset for Tumor Segmentation and Outcome Prediction
Numan Saeed
Salma Hassan
Shahad Hardan
Ahmed Aly
Darya Taratynova
Umair Nawaz
Ufaq Khan
Muhammad Ridzuan
Vincent Andrearczyk
Adrien Depeursinge
Yutong Xie
Thomas Eugene
Raphaël Metz
Mélanie Dore
Gregory Delpon
Vijay Ram Kumar Papineni
Kareem Wahid
Cem Dede
Alaa Mohamed Shawky Ali
Carlos Sjogreen … (see 23 more)
Mohamed Naser
Clifton D. Fuller
Valentin Oreiller
Mario Jreige
John O. Prior
Catherine Cheze Le Rest
Olena Tankyevych
Pierre Decazes
Su Ruan
Stephanie Tanadini-Lang
Hesham Elhalawani
Ronan Abgral
Romain Floch
Kevin Kerleguer
Ulrike Schick
Maelle Mauguen
David Bourhis
Jean-Christophe Leclere
Amandine Sambourg
Arman Rahmim
Mathieu Hatt
Mohammad Yaqub
We describe a publicly available multimodal dataset of annotated Positron Emission Tomography/Computed Tomography (PET/CT) studies for head … (see more)and neck cancer research. The dataset includes 1123 FDG-PET/CT studies from patients with histologically confirmed head and neck cancer, acquired from 10 international medical centers. All examinations consisted of co-registered PET/CT scans with varying acquisition protocols, reflecting real-world clinical diversity across institutions. Primary gross tumor volumes (GTVp) and involved lymph nodes (GTVn) were manually segmented by experienced radiation oncologists and radiologists following standardized guidelines and quality control measures. We provide anonymized NifTi files of all studies, along with expert-annotated segmentation masks, radiotherapy dose distribution for a subset of patients, and comprehensive clinical metadata. This metadata includes TNM staging, HPV status, demographics (age and gender), long-term follow-up outcomes, survival times, censoring indicators, and treatment information. We demonstrate how this dataset can be used for three key clinical tasks: automated tumor segmentation, recurrence-free survival prediction, and HPV status classification, providing benchmark results using state-of-the-art deep learning models, including UNet, SegResNet, and multimodal prognostic frameworks.
Predictive Performance Precision Analysis in Medicine: Identification of low-confidence predictions at patient and profile levels (MED3pa I)
Félix Camirand Lemyre
Jean-François Ethier
Lyna Hiba Chikouche
Ludmila Amriou
Artificial Intelligence models are increasingly used in healthcare, yet global performance metrics can mask variations in reliability across… (see more) individual patients or subgroups with shared attributes, called patient profiles . This study introduces MED3pa, a method that identifies when models are less reliable, allowing clinicians to better assess model limitations. We propose a framework that estimates predictive confidence using three combined approaches: Individualized (IPC), Aggregated (APC), and Mixed Predictive Confidence (MPC). IPC estimates confidence for each patient, APC assesses it across profiles, and MPC combines both. We evaluate our method on four datasets: one simulated, two public, and one private clinical dataset. Metrics by Declaration Rate (MDR) curves show how performance changes when retaining only the most confident predictions, while interpretable decision trees reveal profiles with higher or lower model confidence. We demonstrate our method in internal, temporal, and external validation settings, as well as through a clinical example. In internal validation, limiting predictions to the 93% most confident cases improved sensitivity by 14.3% and the AUC by 5.1%. In the clinical example, MED3pa identified a patient profile with high misclassification risk, demonstrating its potential for safer deployment. By identifying low-confidence predictions, our framework improves model reliability in clinical settings. It can be integrated into decision support systems to help clinicians make more informed decisions. Confidence thresholds help balance model performance with the proportion of patients for whom predictions are considered reliable. Better leveraging confidence in model predictions could improve reliability and trustworthiness, supporting safer and more effective use in healthcare.