arXiv subjects

Sophie Tezenas du Montcel

Publications and source records attributed to Sophie Tezenas du Montcel.

5 recordsLinked to original sources

arXiv2026

Performance uncertainty in medical image analysis: a large-scale investigation of confidence intervals

Performance uncertainty quantification is essential for reliable validation and eventual clinical translation of medical imaging artificial intelligence (AI). Confidence intervals (CIs) play a central role in this process by indicating how precise a reported performance estimate is. Yet, due to the limited amount of work examining CI behavior in medical imaging, the community remains largely unaware of how many diverse CI methods exist and how they behave in specific settings. The purpose of this study is to close this gap. To this end, we conducted a large-scale empirical analysis across a total of 24 segmentation and classification tasks, using 19 trained models per task group, a broad spectrum of commonly used performance metrics, multiple aggregation strategies, and several widely adopted CI methods. Reliability (coverage) and precision (width) of each CI method were estimated across all settings to characterize their dependence on study characteristics. Our analysis revealed five principal findings: 1) the sample size required for reliable CIs varies from a few dozens to several thousands of cases depending on study parameters; 2) CI behavior is strongly affected by the choice of performance metric; 3) aggregation strategy substantially influences the reliability of CIs, e.g. they require more observations for macro than for micro; 4) the machine learning problem (segmentation versus classification) modulates these effects; 5) different CI methods are not equally reliable and precise depending on the use case. Finally, we derived practical implications of this study in the form of a decision tree which shall prove useful to the community. This paves the way for future consensus guidelines on reporting performance uncertainty.

cs.CV↗

arXiv2026

A mixture model for subtype identification in the context of disease progression modeling

The progression of chronic diseases often follows highly variable trajectories, and the underlying factors remain poorly understood. Standard mixed-effects models typically represent inter-patient differences as random deviations around a common reference, which may obscure meaningful subgroups. We propose a probabilistic mixture extension of a mixed effects model, the Disease Course Mapping model, to identify distinct disease progression subtypes within a population. The mixture structure is introduced at the latent individual parameters, enabling clustering based on both temporal and spatial variability in disease trajectories. We evaluated the model through simulation studies to assess classification performance and parameter recovery. Classification accuracy exceeded 90% in simpler scenarios and remained above 80% in the most complex case, with particularly high recall and precision for fast-progressing clusters. Compared to a post hoc classification approach, the proposed model yielded more accurate parameter estimates, smaller biases, lower root mean squared errors, and reduced uncertainty. It also correctly recovered the true three-cluster structure in 93% of the simulations. Finally, we applied the model to a longitudinal cohort of CADASIL patients, identifying two clinically meaningful clusters, differentiating patients with early versus late onset and fast versus slow progression, with clear spatial patterns across motor and memory scores. Overall, this probabilistic mixture framework offers a robust, interpretable approach for clustering patients based on spatiotemporal disease dynamics.

stat.ME↗

arXiv2025

A joint spatiotemporal model for multiple longitudinal markers and competing events

Non-terminal events can represent a meaningful change in a patient's life. Thus, better understanding and predicting their occurrence can bring valuable information to individuals. In a context where longitudinal markers could inform these events, joint models with competing risks have been developed. Their precision relies on a reference time for which disease onset is often used. Nevertheless, chronic diseases have no clear onset, making it difficult to define a precise reference time. We propose a Joint cause-specific Spatiotemporal model to overcome this limitation and to capture a shared latent process, a latent age (temporal aspect), associated with the ordering of the longitudinal outcomes (spatial aspect). First, we validated our model on simulated real-like data. Then, we benchmarked our model with a shared-random-effect joint model on real ALS data using the PRO-ACT dataset. Finally, to show how the model could be used for description tasks, we analysed the impact of sex and onset site on the progression of ALS as well as the initiation of Non-Invasive Ventilation. The Joint cause-specific spatiotemporal model achieved similar performance to the shared random effect joint model while capturing the latent disease age and the impact of the ordering of longitudinal outcomes on the occurrence of the events with fewer parameters. The application study confirmed existing results for the Longitudinal outcomes and showed how to interpret the model. The proposed approach by disentangling a temporal and a spatial aspect of the disease opens the perspective to capture meaningful change in future clinical trials.

stat.ME↗

arXiv2024

Joint model with latent disease age: overcoming the need for reference time

Introduction: Heterogeneity of the progression of neurodegenerative diseases is one of the main challenges faced in developing effective therapies. With the increasing number of large clinical databases, disease progression models have led to a better understanding of this heterogeneity. Nevertheless, these diseases may have no clear onset and biological underlying processes may start before the first symptoms. Such an ill-defined disease reference time is an issue for current joint models, which have proven their effectiveness by combining longitudinal and survival data. Objective In this work, we propose a joint non-linear mixed effect model with a latent disease age, to overcome this need for a precise reference time. Method: To do so, we utilized an existing longitudinal model with a latent disease age as a longitudinal sub-model and associated it with a survival sub-model that estimates a Weibull distribution from the latent disease age. We then validated our model on different simulated scenarios. Finally, we benchmarked our model with a state-of-the-art joint model and reference survival and longitudinal models on simulated and real data in the context of Amyotrophic Lateral Sclerosis (ALS). Results: On real data, our model got significantly better results than the state-of-the-art joint model for absolute bias (4.21(4.41) versus 4.24(4.14)(p-value=1.4e-17)), and mean cumulative AUC for right censored events (0.67(0.07) versus 0.61(0.09)(p-value=1.7e-03)). Conclusion: We showed that our approach is better suited than the state-of-the-art in the context where the reference time is not reliable. This work opens up the perspective to design predictive and personalized therapeutic strategies.

stat.ME↗

arXiv2020

Prevalence and incidence of postpartum depression and environmental factors: the IGEDEPP cohort

Background: IGEDEPP (Interaction of Gene and Environment of Depression during PostPartum) is a prospective multicenter cohort study of 3,310 Caucasian women who gave birth between 2011 and 2016, with follow-up until one year postpartum. The aim of the current study is to describe the cohort and estimate the prevalence and cumulative incidence of early and late postpartum depression (PPD). Methods: Socio-demographic data, personal and family psychiatric history, as well as stressful life events during childhood and pregnancy were evaluated at baseline. Early and late PPD were assessed at 8 weeks and 1 year postpartum respectively, using DSM-5 criteria. Results: The prevalence of early PPD was 8.3% (95%CI 7.3-9.3), and late PPD 12.9% (95%CI 11.5-14.2), resulting in an 8-week cumulative incidence of 8.5% (95%CI 7.4-9.6) and a one-year cumulative incidence of PPD of 18.1% (95%CI: 17.1-19.2). Nearly half of the cohort (N=1571, 47.5%) had a history of at least one psychiatric or addictive disorder, primarily depressive disorder (35%). Almost 300 women in the cohort (9.0%) reported childhood trauma. During pregnancy, 47.7% women experienced a stressful event, 30.2% in the first 8 weeks and 43.9% between 8 weeks and one year postpartum. Nearly one in five women reported at least one stressful postpartum event at 8 weeks. Conclusion: Incident depressive episodes affected nearly one in five women during the first year postpartum. Most women had stressful perinatal events. Further IGEDEPP studies will aim to disentangle the impact of childhood and pregnancy-related stressful events on postpartum mental disorders.

q-bio.QM↗