Thomas Lampert

Federico Zucchi, Yi Xie, Chao Zhang et al.

Adaptive patching is a recent and compelling proposal for time-series Transformers: allocate finer patches where the sequence looks locally informative. This paper asks under what conditions a content-adaptive patching operator should outperform a tuned uniform one. Local heterogeneity alone is not enough: under pointwise forecasting losses, a complex-looking region is not automatically one where finer patching reduces the loss. We model patching as a budgeted bitrate allocation and derive an explicit threshold that a dynamic patching rule must satisfy to beat a well-tuned uniform baseline, then bound the achievable improvement both locally (a quadratic surrogate) and globally (a strong-convexity bound under the model's assumptions). Two structural results follow: without a coupling constraint, scalar local complexity cannot produce a non-uniform optimum under a common loss landscape; and once the backbone is trained to its representation-aware optimum, the alignment gain collapses around a well-tuned uniform patch size. To test these predictions, we run a controlled isolation study on three representative architectures, replacing each adaptive mechanism with a uniform patch-size sweep while keeping the backbone, data, and training protocol fixed. On standard long-horizon forecasting benchmarks, the validation-selected uniform baseline is competitive with the dynamic counterpart, with per-setting effects concentrated near zero and no consistent directional advantage once results are aggregated by dataset. The larger gains we do observe are method- and dataset-specific. Adaptive patching should therefore be evaluated against a tuned uniform baseline; its value depends on whether a cheap and reliable routing signal can identify where finer patches actually reduce forecasting loss.

Jelica Vasiljević, Friedrich Feuerhake, Cédric Wemmert et al.

Virtual stain transfer is a promising area of research in Computational Pathology, which has a great potential to alleviate important limitations when applying deeplearningbased solutions such as lack of annotations and sensitivity to a domain shift. However, in the literature, the majority of virtual staining approaches are trained for a specific staining or stain combination, and their extension to unseen stainings requires the acquisition of additional data and training. In this paper, we propose HistoStarGAN, a unified framework that performs stain transfer between multiple stainings, stain normalisation and stain invariant segmentation, all in one inference of the model. We demonstrate the generalisation abilities of the proposed solution to perform diverse stain transfer and accurate stain invariant segmentation over numerous unseen stainings, which is the first such demonstration in the field. Moreover, the pre-trained HistoStar-GAN model can serve as a synthetic data generator, which paves the way for the use of fully annotated synthetic image data to improve the training of deep learning-based algorithms. To illustrate the capabilities of our approach, as well as the potential risks in the microscopy domain, inspired by applications in natural images, we generated KidneyArtPathology, a fully annotated artificial image dataset for renal pathology.

Zeeshan Nisar, Friedrich Feuerhake, Thomas Lampert

A key challenge in segmentation in digital histopathology is inter- and intra-stain variations as it reduces model performance. Labelling each stain is expensive and time-consuming so methods using stain transfer via CycleGAN, have been developed for training multi-stain segmentation models using labels from a single stain. Nevertheless, CycleGAN tends to introduce noise during translation because of the one-to-many nature of some stain pairs, which conflicts with its cycle consistency loss. To address this, we propose the Domain Shift Aware CycleGAN, which reduces the presence of such noise. Furthermore, we evaluate several advances from the field of machine learning aimed at resolving similar problems and compare their effectiveness against DSA-CycleGAN in the context of multi-stain glomeruli segmentation. Experiments demonstrate that DSA-CycleGAN not only improves segmentation performance in glomeruli segmentation but also outperforms other methods in reducing noise. This is particularly evident when translating between biologically distinct stains. The code is publicly available at https://github.com/zeeshannisar/DSA-CycleGAN.

Zeeshan Nisar, Jelica Vasiljević, Pierre Gançarski et al.

Domain shift in digital histopathology can occur when different stains or scanners are used, during stain translation, etc. A deep neural network trained on source data may not generalise well to data that has undergone some domain shift. An important step towards being robust to domain shift is the ability to detect and measure it. This article demonstrates that the PixelCNN and domain shift metric can be used to detect and quantify domain shift in digital histopathology, and they demonstrate a strong correlation with generalisation performance. These findings pave the way for a mechanism to infer the average performance of a model (trained on source data) on unseen and unlabelled target data.

Zeeshan Nisar, Friedrich Feuerhake, Thomas Lampert

Semantic segmentation under domain shift remains a fundamental challenge in computer vision, particularly when labelled training data is scarce. This challenge is particularly exemplified in histopathology image analysis, where the same tissue structures must be segmented across images captured under different imaging conditions (stains), each representing a distinct visual domain. Traditional deep learning methods like UNet require extensive labels, which is both costly and time-consuming, particularly when dealing with multiple domains (or stains). To mitigate this, various unsupervised domain adaptation based methods such as UDAGAN have been proposed, which reduce the need for labels by requiring only one (source) stain to be labelled. Nonetheless, obtaining source stain labels can still be challenging. This article shows that through self-supervised pre-training -- including SimCLR, BYOL, and a novel approach, HR-CS-CO -- the performance of these segmentation methods (UNet, and UDAGAN) can be retained even with 95% fewer labels. Notably, with self-supervised pre-training and using only 5% labels, the performance drops are minimal: 5.9% for UNet and 6.2% for UDAGAN, averaged over all stains, compared to their respective fully supervised counterparts (without pre-training, using 100% labels). Furthermore, these findings are shown to generalise beyond their training distribution to public benchmark datasets. Implementations and pre-trained models are publicly available \href{https://github.com/zeeshannisar/resource-effecient-multi-stain-kidney-glomeruli-segmentation.git}{online}.

Niklas Kormann, Masoud Ramuz, Zeeshan Nisar et al.

Graph Neural Networks (GNNs) have recently been found to excel in histopathology. However, an important histopathological task, where GNNs have not been extensively explored, is the classification of glomeruli health as an important indicator in nephropathology. This task presents unique difficulties, particularly for the graph construction, i.e., the identification of nodes, edges, and informative features. In this work, we propose a pipeline composed of different traditional and machine learning-based computer vision techniques to identify nodes, edges, and their corresponding features to form a heterogeneous graph. We then proceed to propose a novel heterogeneous GNN architecture for glomeruli classification, called HIEGNet, that integrates both glomeruli and their surrounding immune cells. Hence, HIEGNet is able to consider the immune environment of each glomerulus in its classification. Our HIEGNet was trained and tested on a dataset of Whole Slide Images from kidney transplant patients. Experimental results demonstrate that HIEGNet outperforms several baseline models and generalises best between patients among all baseline models. Our implementation is publicly available at https://github.com/nklsKrmnn/HIEGNet.git.

Federico Zucchi, Thomas Lampert

Multivariate time-series classification is pivotal in domains ranging from wearable sensing to biomedical monitoring. Despite recent advances, Transformer- and CNN-based models often remain computationally heavy, offer limited frequency diversity, and require extensive parameter budgets. We propose PRISM (Per-channel Resolution-Informed Symmetric Module), a convolutional-based feature extractor that applies symmetric finite-impulse-response (FIR) filters at multiple temporal scales, independently per channel. This multi-resolution, per-channel design yields highly frequency-selective embeddings without any inter-channel convolutions, greatly reducing model size and complexity. Across human-activity, sleep-stage and biomedical benchmarks, PRISM, paired with lightweight classification heads, matches or outperforms leading CNN and Transformer baselines, while using roughly an order of magnitude fewer parameters and FLOPs. By uniting classical signal processing insights with modern deep learning, PRISM offers an accurate, resource-efficient solution for multivariate time-series classification.

Jelica Vasiljević, Friedrich Feuerhake, Cédric Wemmert et al.

It has been shown that unpaired image-to-image translation methods constrained by cycle-consistency hide the information necessary for accurate input reconstruction as imperceptible noise. We demonstrate that, when applied to histopathology data, this hidden noise appears to be related to stain specific features and show that this is the case with two immunohistochemical stainings during translation to Periodic acid- Schiff (PAS), a histochemical staining method commonly applied in renal pathology. Moreover, by perturbing this hidden information, the translation models produce different, plausible outputs. We demonstrate that this property can be used as an augmentation method which, in a case of supervised glomeruli segmentation, leads to improved performance.

Jelica Vasiljević, Friedrich Feuerhake, Cédric Wemmert et al.

The application of supervised deep learning methods in digital pathology is limited due to their sensitivity to domain shift. Digital Pathology is an area prone to high variability due to many sources, including the common practice of evaluating several consecutive tissue sections stained with different staining protocols. Obtaining labels for each stain is very expensive and time consuming as it requires a high level of domain knowledge. In this article, we propose an unsupervised augmentation approach based on adversarial image-to-image translation, which facilitates the training of stain invariant supervised convolutional neural networks. By training the network on one commonly used staining modality and applying it to images that include corresponding, but differently stained, tissue structures, the presented method demonstrates significant improvements over other approaches. These benefits are illustrated in the problem of glomeruli segmentation in seven different staining modalities (PAS, Jones H&E, CD68, Sirius Red, CD34, H&E and CD3) and analysis of the learned representations demonstrate their stain invariance.

Odyssee Merveille, Thomas Lampert, Jessica Schmitz et al.

Objective: This article presents an automatic image processing framework to extract quantitative high-level information describing the micro-environment of glomeruli in consecutive whole slide images (WSIs) processed with different staining modalities of patients with chronic kidney rejection after kidney transplantation. Methods: This four-step framework consists of: 1) approximate rigid registration, 2) cell and anatomical structure segmentation 3) fusion of information from different stainings using a newly developed registration algorithm 4) feature extraction. Results: Each step of the framework is validated independently both quantitatively and qualitatively by pathologists. An illustration of the different types of features that can be extracted is presented. Conclusion: The proposed generic framework allows for the analysis of the micro-environment surrounding large structures that can be segmented (either manually or automatically). It is independent of the segmentation approach and is therefore applicable to a variety of biomedical research questions. Significance: Chronic tissue remodelling processes after kidney transplantation can result in interstitial fibrosis and tubular atrophy (IFTA) and glomerulosclerosis. This pipeline provides tools to quantitatively analyse, in the same spatial context, information from different consecutive WSIs and help researchers understand the complex underlying mechanisms leading to IFTA and glomerulosclerosis.

Thomas Lampert, Odyssée Merveille, Jessica Schmitz et al.

An important part of Digital Pathology is the analysis of multiple digitised whole slide images from differently stained tissue sections. It is common practice to mount consecutive sections containing corresponding microscopic structures on glass slides, and to stain them differently to highlight specific tissue components. These multiple staining modalities result in very different images but include a significant amount of consistent image information. Deep learning approaches have recently been proposed to analyse these images in order to automatically identify objects of interest for pathologists. These supervised approaches require a vast amount of annotations, which are difficult and expensive to acquire---a problem that is multiplied with multiple stainings. This article presents several training strategies that make progress towards stain invariant networks. By training the network on one commonly used staining modality and applying it to images that include corresponding but differently stained tissue structures, the presented unsupervised strategies demonstrate significant improvements over standard training strategies.

Pedro Chahuara, Thomas Lampert, Pierre Gancarski

Presented herein is a novel model for similar question ranking within collaborative question answer platforms. The presented approach integrates a regression stage to relate topics derived from questions to those derived from question-answer pairs. This helps to avoid problems caused by the differences in vocabulary used within questions and answers, and the tendency for questions to be shorter than answers. The performance of the model is shown to outperform translation methods and topic modelling (without regression) on several real-world datasets.