arXiv subjects

Sarah Müller

Publications and source records attributed to Sarah Müller.

9 recordsLinked to original sources

arXiv2026

Ordinal Diffusion Models for Color Fundus Images

Generative image models such as diffusion models can improve performance on clinically relevant tasks by offering deep learning models supplementary training data. However, most conditional diffusion models treat disease stages as independent classes, ignoring the continuous nature of disease progression. This mismatch is problematic in medical imaging because continuous pathological processes are typically only observed through coarse, discrete but ordered labels as in ophthalmology for diabetic retinopathy (DR). We propose an ordinal latent diffusion model for generating color fundus images that explicitly incorporates the ordered structure of DR severity into the generation process. Instead of categorical conditioning, we used a scalar disease representation, enabling a smooth transition between adjacent stages. We evaluated our approach using visual realism metrics and classification-based clinical consistency analysis on the EyePACS dataset. Compared to a standard conditional diffusion model, our model reduced the Fr\'echet inception distance for four of the five DR stages and increased the quadratic weighted $\kappa$ from 0.79 to 0.87. Furthermore, interpolation experiments showed that the model captured a continuous spectrum of disease progression learned from ordered, coarse class labels. Code available at https://github.com/berenslab/OrdinalDiffusionModels.

cs.CV↗

arXiv2026

Mitigating Shortcut Learning via Feature Disentanglement in Medical Imaging: A Benchmark Study

Although deep learning models in medical imaging often achieve excellent classification performance, they can rely on shortcut learning, exploiting spurious correlations or confounding factors that are not causally related to the target task. This poses risks in clinical settings, where models must generalize across institutions, populations, and acquisition conditions. Feature disentanglement is a promising approach to mitigate shortcut learning by separating task-relevant information from confounder-related features in latent representations. In this study, we systematically evaluated feature disentanglement methods for mitigating shortcuts in medical imaging, including adversarial learning and latent space splitting based on dependence minimization. We assessed classification performance and disentanglement quality using latent space analyses across one artificial and two medical datasets with natural and synthetic confounders. We also examined robustness under varying levels of confounding and compared computational efficiency across methods. We found that shortcut mitigation methods improved classification performance under strong spurious correlations during training. Latent space analyses revealed differences in representation quality not captured by classification metrics, highlighting the strengths and limitations of each method. Model reliance on shortcuts depended on the degree of confounding in the training data. The best-performing models combine data-centric rebalancing with model-centric disentanglement, achieving stronger and more robust shortcut mitigation than rebalancing alone while maintaining similar computational efficiency. The project code is publicly available at https://github.com/berenslab/medical-shortcut-mitigation.

cs.CV↗

arXiv2025

Learning Disease State from Noisy Ordinal Disease Progression Labels

Learning from noisy ordinal labels is a key challenge in medical imaging. In this work, we ask whether ordinal disease progression labels (better, worse, or stable) can be used to learn a representation allowing to classify disease state. For neovascular age-related macular degeneration (nAMD), we cast the problem of modeling disease progression between medical visits as a classification task with ordinal ranks. To enhance generalization, we tailor our model to the problem setting by (1) independent image encoding, (2) antisymmetric logit space equivariance, and (3) ordinal scale awareness. In addition, we address label noise by learning an uncertainty estimate for loss re-weighting. Our approach learns an interpretable disease representation enabling strong few-shot performance for the related task of nAMD activity classification from single images, despite being trained only on image pairs with ordinal disease progression labels.

cs.CV↗

arXiv2025

Disentangling representations of retinal images with generative models

Retinal fundus images play a crucial role in the early detection of eye diseases. However, the impact of technical factors on these images can pose challenges for reliable AI applications in ophthalmology. For example, large fundus cohorts are often confounded by factors like camera type, bearing the risk of learning shortcuts rather than the causal relationships behind the image generation process. Here, we introduce a population model for retinal fundus images that effectively disentangles patient attributes from camera effects, enabling controllable and highly realistic image generation. To achieve this, we propose a disentanglement loss based on distance correlation. Through qualitative and quantitative analyses, we show that our models encode desired information in disentangled subspaces and enable controllable image generation based on the learned subspaces, demonstrating the effectiveness of our disentanglement loss. The project's code is publicly available: https://github.com/berenslab/disentangling-retinal-images.

cs.CV↗

arXiv2024

Benchmarking Dependence Measures to Prevent Shortcut Learning in Medical Imaging

Medical imaging cohorts are often confounded by factors such as acquisition devices, hospital sites, patient backgrounds, and many more. As a result, deep learning models tend to learn spurious correlations instead of causally related features, limiting their generalizability to new and unseen data. This problem can be addressed by minimizing dependence measures between intermediate representations of task-related and non-task-related variables. These measures include mutual information, distance correlation, and the performance of adversarial classifiers. Here, we benchmark such dependence measures for the task of preventing shortcut learning. We study a simplified setting using Morpho-MNIST and a medical imaging task with CheXpert chest radiographs. Our results provide insights into how to mitigate confounding factors in medical imaging.

cs.CV↗

arXiv2024

Most discriminative stimuli for functional cell type clustering

Identifying cell types and understanding their functional properties is crucial for unraveling the mechanisms underlying perception and cognition. In the retina, functional types can be identified by carefully selected stimuli, but this requires expert domain knowledge and biases the procedure towards previously known cell types. In the visual cortex, it is still unknown what functional types exist and how to identify them. Thus, for unbiased identification of the functional cell types in retina and visual cortex, new approaches are needed. Here we propose an optimization-based clustering approach using deep predictive models to obtain functional clusters of neurons using Most Discriminative Stimuli (MDS). Our approach alternates between stimulus optimization with cluster reassignment akin to an expectation-maximization algorithm. The algorithm recovers functional clusters in mouse retina, marmoset retina and macaque visual area V4. This demonstrates that our approach can successfully find discriminative stimuli across species, stages of the visual system and recording techniques. The resulting most discriminative stimuli can be used to assign functional cell types fast and on the fly, without the need to train complex predictive models or show a large natural scene dataset, paving the way for experiments that were previously limited by experimental time. Crucially, MDS are interpretable: they visualize the distinctive stimulus patterns that most unambiguously identify a specific type of neuron.

q-bio.NC↗

arXiv2021

Local policy search with Bayesian optimization

Reinforcement learning (RL) aims to find an optimal policy by interaction with an environment. Consequently, learning complex behavior requires a vast number of samples, which can be prohibitive in practice. Nevertheless, instead of systematically reasoning and actively choosing informative samples, policy gradients for local search are often obtained from random perturbations. These random samples yield high variance estimates and hence are sub-optimal in terms of sample complexity. Actively selecting informative samples is at the core of Bayesian optimization, which constructs a probabilistic surrogate of the objective from past samples to reason about informative subsequent ones. In this paper, we propose to join both worlds. We develop an algorithm utilizing a probabilistic model of the objective function and its gradient. Based on the model, the algorithm decides where to query a noisy zeroth-order oracle to improve the gradient estimates. The resulting algorithm is a novel type of policy search method, which we compare to existing black-box algorithms. The comparison reveals improved sample complexity and reduced variance in extensive empirical evaluations on synthetic objectives. Further, we highlight the benefits of active sampling on popular RL benchmarks.

cs.LG↗

arXiv2015

Sum Rules of Charm CP Asymmetries beyond the SU(3)$_F$ Limit

We find new sum rules between direct CP asymmetries in $D$ meson decays with coefficients that can be determined from a global fit to branching ratio data. Our sum rules eliminate the penguin topologies $P$ and $PA$, which cannot be determined from branching ratios. In this way we can make predictions about direct CP asymmetries in the standard model without ad hoc assumptions on the sizes of penguin diagrams. We consistently include first-order SU(3)$_F$ breaking in the topological amplitudes extracted from the branching ratios. By confronting our sum rules with future precise data from LHCb and Belle~II one will identify or constrain new-physics contributions to $P$ or $PA$. The first sum rule correlates the CP asymmetries $a_{CP}^{\mathrm{dir}}$ in $D^0\to K^+K^-$, $D^0\to π^+π^-$, and $D^0\to π^0π^0$. We study the region of the $a_{CP}^{\mathrm{dir}}(D^0\to π^+π^-)$--$a_{CP}^{\mathrm{dir}} (D^0\to π^0π^0)$ plane allowed by current data and find that our sum rule excludes more than half of the allowed region at 95\% C.L. Our second sum rule correlates the direct CP asymmetries in $D^+\rightarrow \bar{K}^0 K^+$, $D_s^+\rightarrow K^0 π^+$, and $D_s^+\rightarrow K^+π^0$.

hep-ph↗

arXiv2015

Topological amplitudes in $D$ decays to two pseudoscalars: a global analysis with linear $SU(3)_F$ breaking

We study decays of $D^0$, $D^+$, and $D_s^+$ mesons into two pseudoscalar mesons by expressing the decay amplitudes in terms of topological amplitudes. Including consistently SU(3)$_F$ breaking to linear order, we show how the topological-amplitude decomposition can be mapped onto the standard expansion using reduced amplitudes characterized by SU(3) representations. The tree and annihilation amplitudes can be calculated in factorization up to corrections which are quadratic in the color-counting parameter $1/N_c$. We find new sum rules connecting $D^+\rightarrow K_SK^+$, $D_s^+\rightarrow K_Sπ^+$ and $D^+\rightarrow K^+π^0$, which test the quality of the $1/N_c$ expansion. Subsequently, we determine the topological amplitudes in a global fit to the data, taking the statistical correlations among the various measurements into account. We carry out likelihood ratio tests in order to quantify the role of specific topological contributions. While the SU(3)$_F$ limit is excluded with a significance of more than five standard deviations, a good fit (with $Δχ^2 <1$) can be obtained with less than $28\%$ of SU(3)$_F$ breaking in the decay amplitudes. The magnitude of the penguin amplitude $P_{\mathrm{break}}$, which probes the Glashow Iliopoulos Maiani (GIM) mechanism, is consistent with zero; the hypothesis $P_{\mathrm{break}}=0$ is rejected with a significance of just $0.7σ$. We obtain the Standard-Model correlation between $\mathcal{B}(D^0\rightarrow K_L π^0)$ and $\mathcal{B}(D^0\rightarrow K_Sπ^0)$, which probes doubly Cabibbo-suppressed amplitudes, and find that $\mathcal{B}(D^0\rightarrow K_L π^0)<\mathcal{B}(D^0\rightarrow K_Sπ^0)$ holds with a significance of more than $4σ$. We finally predict $\mathcal{B}(D_s^+\rightarrow K_L K^+)= 0.012^{+0.006}_{-0.002}$ at $3σ$ CL.

hep-ph↗