Marc Aubreville

Jingna Qiu, Frauke Wilm, Mathias Öttl et al.

The process of annotating histological gigapixel-sized whole slide images (WSIs) at the pixel level for the purpose of training a supervised segmentation model is time-consuming. Region-based active learning (AL) involves training the model on a limited number of annotated image regions instead of requesting annotations of the entire images. These annotation regions are iteratively selected, with the goal of optimizing model performance while minimizing the annotated area. The standard method for region selection evaluates the informativeness of all square regions of a specified size and then selects a specific quantity of the most informative regions. We find that the efficiency of this method highly depends on the choice of AL step size (i.e., the combination of region size and the number of selected regions per WSI), and a suboptimal AL step size can result in redundant annotation requests or inflated computation costs. This paper introduces a novel technique for selecting annotation regions adaptively, mitigating the reliance on this AL hyperparameter. Specifically, we dynamically determine each region by first identifying an informative area and then detecting its optimal bounding box, as opposed to selecting regions of a uniform predefined shape and size as in the standard method. We evaluate our method using the task of breast cancer metastases segmentation on the public CAMELYON16 dataset and show that it consistently achieves higher sampling efficiency than the standard method across various AL step sizes. With only 2.6\% of tissue area annotated, we achieve full annotation performance and thereby substantially reduce the costs of annotating a WSI dataset. The source code is available at https://github.com/DeepMicroscopy/AdaptiveRegionSelection.

Matthias Eisenmann, Annika Reinke, Vivienn Weru et al.

International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi-center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), image preprocessing (97%), data curation (79%), and postprocessing (66%). The "typical" lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.

Marc Aubreville, Nikolas Stathonikos, Christof A. Bertram et al.

The density of mitotic figures within tumor tissue is known to be highly correlated with tumor proliferation and thus is an important marker in tumor grading. Recognition of mitotic figures by pathologists is known to be subject to a strong inter-rater bias, which limits the prognostic value. State-of-the-art deep learning methods can support the expert in this assessment but are known to strongly deteriorate when applied in a different clinical environment than was used for training. One decisive component in the underlying domain shift has been identified as the variability caused by using different whole slide scanners. The goal of the MICCAI MIDOG 2021 challenge has been to propose and evaluate methods that counter this domain shift and derive scanner-agnostic mitosis detection algorithms. The challenge used a training set of 200 cases, split across four scanning systems. As a test set, an additional 100 cases split across four scanning systems, including two previously unseen scanners, were given. The best approaches performed on an expert level, with the winning algorithm yielding an F_1 score of 0.748 (CI95: 0.704-0.781). In this paper, we evaluate and compare the approaches that were submitted to the challenge and identify methodological factors contributing to better performance.

Marc Aubreville, Nikolas Stathonikos, Taryn A. Donovan et al.

Recognition of mitotic figures in histologic tumor specimens is highly relevant to patient outcome assessment. This task is challenging for algorithms and human experts alike, with deterioration of algorithmic performance under shifts in image representations. Considerable covariate shifts occur when assessment is performed on different tumor types, images are acquired using different digitization devices, or specimens are produced in different laboratories. This observation motivated the inception of the 2022 challenge on MItosis Domain Generalization (MIDOG 2022). The challenge provided annotated histologic tumor images from six different domains and evaluated the algorithmic approaches for mitotic figure detection provided by nine challenge participants on ten independent domains. Ground truth for mitotic figure detection was established in two ways: a three-expert consensus and an independent, immunohistochemistry-assisted set of labels. This work represents an overview of the challenge tasks, the algorithmic strategies employed by the participants, and potential factors contributing to their success. With an $F_1$ score of 0.764 for the top-performing team, we summarize that domain generalization across various tumor domains is possible with today's deep learning-based recognition pipelines. However, we also found that domain characteristics not present in the training set (feline as new species, spindle cell shape as new morphology and a new scanner) led to small but significant decreases in performance. When assessed against the immunohistochemistry-assisted reference standard, all methods resulted in reduced recall scores, but with only minor changes in the order of participants in the ranking.

Jingna Qiu, Marc Aubreville, Frauke Wilm et al.

Acquiring annotations for whole slide images (WSIs)-based deep learning tasks, such as creating tissue segmentation masks or detecting mitotic figures, is a laborious process due to the extensive image size and the significant manual work involved in the annotation. This paper focuses on identifying and annotating specific image regions that optimize model training, given a limited annotation budget. While random sampling helps capture data variance by collecting annotation regions throughout the WSIs, insufficient data curation may result in an inadequate representation of minority classes. Recent studies proposed diversity sampling to select a set of regions that maximally represent unique characteristics of the WSIs. This is done by pretraining on unlabeled data through self-supervised learning and then clustering all regions in the latent space. However, establishing the optimal number of clusters can be difficult and not all clusters are task-relevant. This paper presents prototype sampling, a new method for annotation region selection. It discovers regions exhibiting typical characteristics of each task-specific class. The process entails recognizing class prototypes from extensive histopathology image-caption databases and detecting unlabeled image regions that resemble these prototypes. Our results show that prototype sampling is more effective than random and diversity sampling in identifying annotation regions with valuable training information, resulting in improved model performance in semantic segmentation and mitotic figure detection tasks. Code is available at https://github.com/DeepMicroscopy/Prototype-sampling.

Jingna Qiu, Frauke Wilm, Mathias Öttl et al.

Active learning improves annotation efficiency by selecting the most informative samples for annotation and model training. While most prior work has focused on selecting informative images for classification tasks, we investigate the more challenging setting of dense prediction, where annotations are more costly and time-intensive, especially in medical imaging. Region-level annotation has been shown to be more efficient than image-level annotation for these tasks. However, existing methods for representative annotation region selection suffer from high computational and memory costs, irrelevant region choices, and heavy reliance on uncertainty sampling. We propose decomposition sampling (DECOMP), a new active learning sampling strategy that addresses these limitations. It enhances annotation diversity by decomposing images into class-specific components using pseudo-labels and sampling regions from each class. Class-wise predictive confidence further guides the sampling process, ensuring that difficult classes receive additional annotations. Across ROI classification, 2-D segmentation, and 3-D segmentation, DECOMP consistently surpasses baseline methods by better sampling minority-class regions and boosting performance on these challenging classes. Code is in https://github.com/JingnaQiu/DECOMP.git.

Sweta Banerjee, Timo Gosch, Sara Hester et al.

The annotation of large scale histopathology image datasets remains a major bottleneck in developing robust deep learning models for clinically relevant tasks, such as mitotic figure classification. Folder-based annotation workflows are usually slow, fatiguing, and difficult to scale. To address these challenges, we introduce SWipeable ANnotations (SWAN), an open-source, MIT-licensed web application that enables intuitive image patch classification using a swiping gesture. SWAN supports both desktop and mobile platforms, offers real-time metadata capture, and allows flexible mapping of swipe gestures to class labels. In a pilot study with four pathologists annotating 600 mitotic figure image patches, we compared SWAN against a traditional folder-sorting workflow. SWAN enabled rapid annotations with pairwise percent agreement ranging from 86.52% to 93.68% (Cohen's Kappa = 0.61-0.80), while for the folder-based method, the pairwise percent agreement ranged from 86.98% to 91.32% (Cohen's Kappa = 0.63-0.75) for the task of classifying atypical versus normal mitotic figures, demonstrating high consistency between annotators and comparable performance. Participants rated the tool as highly usable and appreciated the ability to annotate on mobile devices. These results suggest that SWAN can accelerate image annotation while maintaining annotation quality, offering a scalable and user-friendly alternative to conventional workflows.

Andreas Haghofer, Eda Parlak, Alexander Bartel et al.

Variation in nuclear size and shape is an important criterion of malignancy for many tumor types; however, categorical estimates by pathologists have poor reproducibility. Measurements of nuclear characteristics (morphometry) can improve reproducibility, but manual methods are time consuming. The aim of this study was to explore the limitations of estimates and develop alternative morphometric solutions for canine cutaneous mast cell tumors (ccMCT). We assessed the following nuclear evaluation methods for measurement accuracy, reproducibility, and prognostic utility: 1) anisokaryosis (karyomegaly) estimates by 11 pathologists; 2) gold standard manual morphometry of at least 100 nuclei; 3) practicable manual morphometry with stratified sampling of 12 nuclei by 9 pathologists; and 4) automated morphometry using a deep learning-based segmentation algorithm. The study dataset comprised 96 ccMCT with available outcome information. The study dataset comprised 96 ccMCT with available outcome information. Inter-rater reproducibility of karyomegaly estimates was low ($κ$ = 0.226), while it was good (ICC = 0.654) for practicable morphometry of the standard deviation (SD) of nuclear size. As compared to gold standard manual morphometry (AUC = 0.839, 95% CI: 0.701 - 0.977), the prognostic value (tumor-specific survival) of SDs of nuclear area for practicable manual morphometry (12 nuclei) and automated morphometry were high with an area under the ROC curve (AUC) of 0.868 (95% CI: 0.737 - 0.991) and 0.943 (95% CI: 0.889 - 0.996), respectively. This study supports the use of manual morphometry with stratified sampling of 12 nuclei and algorithmic morphometry to overcome the poor reproducibility of estimates.

Frauke Wilm, Marco Fragoso, Christof A. Bertram et al.

In histopathology, scanner-induced domain shifts are known to impede the performance of trained neural networks when tested on unseen data. Multi-domain pre-training or dedicated domain-generalization techniques can help to develop domain-agnostic algorithms. For this, multi-scanner datasets with a high variety of slide scanning systems are highly desirable. We present a publicly available multi-scanner dataset of canine cutaneous squamous cell carcinoma histopathology images, composed of 44 samples digitized with five slide scanners. This dataset provides local correspondences between images and thereby isolates the scanner-induced domain shift from other inherent, e.g. morphology-induced domain shifts. To highlight scanner differences, we present a detailed evaluation of color distributions, sharpness, and contrast of the individual scanner subsets. Additionally, to quantify the inherent scanner-induced domain shift, we train a tumor segmentation network on each scanner subset and evaluate the performance both in- and cross-domain. We achieve a class-averaged in-domain intersection over union coefficient of up to 0.86 and observe a cross-domain performance decrease of up to 0.38, which confirms the inherent domain shift of the presented dataset and its negative impact on the performance of deep neural networks.

Frauke Wilm, Marco Fragoso, Christof A. Bertram et al.

Computer-aided systems in histopathology are often challenged by various sources of domain shift that impact the performance of these algorithms considerably. We investigated the potential of using self-supervised pre-training to overcome scanner-induced domain shifts for the downstream task of tumor segmentation. For this, we present the Barlow Triplets to learn scanner-invariant representations from a multi-scanner dataset with local image correspondences. We show that self-supervised pre-training successfully aligned different scanner representations, which, interestingly only results in a limited benefit for our downstream task. We thereby provide insights into the influence of scanner characteristics for downstream applications and contribute to a better understanding of why established self-supervised methods have not yet shown the same success on histopathology data as they have for natural images.

Hassan Keshvarikhojasteh, Marc Aubreville, Christof A. Bertram et al.

Machine learning models have become integral to many fields, but their reliability, defined as producing dependable, trustworthy, and domain-consistent predictions, remains a critical concern. Multiple Instance Learning (MIL) models designed for Whole Slide Image (WSI) classification in computational pathology are rarely evaluated in terms of reliability, leaving a key gap in understanding their suitability for high-stakes applications like clinical decision-making. In this paper, we address this gap by introducing three quantitative metrics for reliability assessment and applying them to several widely used MIL architectures across three region-wise annotated pathology datasets. Our findings indicate that the mean pooling instance (MEAN-POOL-INS)model demonstrates superior reliability compared to other networks, despite its simple architectural design and computational efficiency. These findings underscore the need of reliability evaluation alongside predictive performance in MIL models and establish MEAN-POOL-INS as a strong, trustworthy baseline for future research.

Jonathan Ganz, Karoline Lipnik, Jonas Ammeling et al.

Nucleolar organizer regions (NORs) are parts of the DNA that are involved in RNA transcription. Due to the silver affinity of associated proteins, argyrophilic NORs (AgNORs) can be visualized using silver-based staining. The average number of AgNORs per nucleus has been shown to be a prognostic factor for predicting the outcome of many tumors. Since manual detection of AgNORs is laborious, automation is of high interest. We present a deep learning-based pipeline for automatically determining the AgNOR-score from histopathological sections. An additional annotation experiment was conducted with six pathologists to provide an independent performance evaluation of our approach. Across all raters and images, we found a mean squared error of 0.054 between the AgNOR- scores of the experts and those of the model, indicating that our approach offers performance comparable to humans.

Pablo Pernias, Dominic Rampas, Mats L. Richter et al.

We introduce Würstchen, a novel architecture for text-to-image synthesis that combines competitive performance with unprecedented cost-effectiveness for large-scale text-to-image diffusion models. A key contribution of our work is to develop a latent diffusion technique in which we learn a detailed but extremely compact semantic image representation used to guide the diffusion process. This highly compressed representation of an image provides much more detailed guidance compared to latent representations of language and this significantly reduces the computational requirements to achieve state-of-the-art results. Our approach also improves the quality of text-conditioned image generation based on our user preference study. The training requirements of our approach consists of 24,602 A100-GPU hours - compared to Stable Diffusion 2.1's 200,000 GPU hours. Our approach also requires less training data to achieve these results. Furthermore, our compact latent representations allows us to perform inference over twice as fast, slashing the usual costs and carbon footprint of a state-of-the-art (SOTA) diffusion model significantly, without compromising the end performance. In a broader comparison against SOTA models our approach is substantially more efficient and compares favorably in terms of image quality. We believe that this work motivates more emphasis on the prioritization of both performance and computational accessibility.

Marc Aubreville, Jonathan Ganz, Jonas Ammeling et al.

Mitotic activity is key for the assessment of malignancy in many tumors. Moreover, it has been demonstrated that the proportion of abnormal mitosis to normal mitosis is of prognostic significance. Atypical mitotic figures (MF) can be identified morphologically as having segregation abnormalities of the chromatids. In this work, we perform, for the first time, automatic subtyping of mitotic figures into normal and atypical categories according to characteristic morphological appearances of the different phases of mitosis. Using the publicly available MIDOG21 and TUPAC16 breast cancer mitosis datasets, two experts blindly subtyped mitotic figures into five morphological categories. Further, we set up a state-of-the-art object detection pipeline extending the anchor-free FCOS approach with a gated hierarchical subclassification branch. Our labeling experiment indicated that subtyping of mitotic figures is a challenging task and prone to inter-rater disagreement, which we found in 24.89% of MF. Using the more diverse MIDOG21 dataset for training and TUPAC16 for testing, we reached a mean overall average precision score of 0.552, a ROC AUC score of 0.833 for atypical/normal MF and a mean class-averaged ROC-AUC score of 0.977 for discriminating the different phases of cells undergoing mitosis.

Marc Aubreville, Zhaoya Pan, Matti Sievert et al.

The surgical removal of head and neck tumors requires safe margins, which are usually confirmed intraoperatively by means of frozen sections. This method is, in itself, an oversampling procedure, which has a relatively low sensitivity compared to the definitive tissue analysis on paraffin-embedded sections. Confocal laser endomicroscopy (CLE) is an in-vivo imaging technique that has shown its potential in the live optical biopsy of tissue. An automated analysis of this notoriously difficult to interpret modality would help surgeons. However, the images of CLE show a wide variability of patterns, caused both by individual factors but also, and most strongly, by the anatomical structures of the imaged tissue, making it a challenging pattern recognition task. In this work, we evaluate four popular few shot learning (FSL) methods towards their capability of generalizing to unseen anatomical domains in CLE images. We evaluate this on images of sinunasal tumors (SNT) from five patients and on images of the vocal folds (VF) from 11 patients using a cross-validation scheme. The best respective approach reached a median accuracy of 79.6% on the rather homogeneous VF dataset, but only of 61.6% for the highly diverse SNT dataset. Our results indicate that FSL on CLE images is viable, but strongly affected by the number of patients, as well as the diversity of anatomical patterns.

Jonas Ammeling, Lars-Henning Schmidt, Jonathan Ganz et al.

Attention-based multiple instance learning (AMIL) algorithms have proven to be successful in utilizing gigapixel whole-slide images (WSIs) for a variety of different computational pathology tasks such as outcome prediction and cancer subtyping problems. We extended an AMIL approach to the task of survival prediction by utilizing the classical Cox partial likelihood as a loss function, converting the AMIL model into a nonlinear proportional hazards model. We applied the model to tissue microarray (TMA) slides of 330 lung cancer patients. The results show that AMIL approaches can handle very small amounts of tissue from a TMA and reach similar C-index performance compared to established survival prediction methods trained with highly discriminative clinical factors such as age, cancer grade, and cancer stage

Christof A. Bertram, Jonas Ammeling, Alexander Bartel et al.

Deep learning-based automated image analysis (DL-AIA) has been shown to outperform trained pathologists in tasks related to feature quantification. Related to these capacities the use of DL-AIA tools is currently extending from proof-of-principle studies to routine applications such as patient samples (diagnostic pathology), regulatory safety assessment (toxicologic pathology), and recurrent research tasks. To ensure that DL-AIA applications are safe and reliable, it is critical to conduct a thorough and objective generalization performance assessment (i.e., the ability of the algorithm to accurately predict patterns of interest) and possibly evaluate model robustness (i.e., the algorithm's capacity to maintain predictive accuracy on images from different sources). In this article, we review the practices for performance assessment in veterinary pathology publications by which two approaches were identified: 1) Exclusive visual performance control (i.e. eyeballing of algorithmic predictions) plus validation of the models application utilizing secondary performance indices, and 2) Statistical performance control (alongside the other methods), which requires a dataset creation and separation of an hold-out test set prior to model training. This article compares the strengths and weaknesses of statistical and visual performance control methods. Furthermore, we discuss relevant considerations for rigorous statistical performance evaluation including metric selection, test dataset image composition, ground truth label quality, resampling methods such as bootstrapping, statistical comparison of multiple models, and evaluation of model stability. It is our conclusion that visual and statistical evaluation have complementary strength and a combination of both provides the greatest insight into the DL model's performance and sources of error.

Emely Rosbach, Jonas Ammeling, Jonathan Ganz et al.

Artificial intelligence (AI)-driven decision support systems can improve diagnostic accuracy and efficiency in computational pathology. However, collaboration between human experts and AI may introduce cognitive biases such as automation and anchoring bias, where users adopt system predictions blindly or are disproportionately influenced by AI advice, even when inaccurate. These effects may be amplified under time pressure, common in routine pathology, or shaped by individual user characteristics. We conducted an online experiment in which pathology experts (n = 28) estimated tumor cell percentages: once independently and once with AI support. A subset of estimations in each condition was performed under time strain. Overall, AI assistance improved diagnostic performance but introduced a 7% automation bias rate, defined as accepted negative consultations where previously correct independent judgments were overturned by incorrect AI advice. While time pressure did not increase the frequency of automation bias, it appeared to intensify its severity, reflected in stronger performance declines associated with increased AI reliance under cognitive load. A linear mixed-effects model (LMM) simulating weighted averaging showed a statistically significant positive coefficient for AI advice, indicating moderate anchoring on system output. This effect increased under time pressure, suggesting anchoring bias becomes more pronounced when cognitive resources are limited. A second LMM assessing automation reliance, a proxy for automation and anchoring bias, showed that professional experience and self-efficacy were associated with lower dependence on AI, whereas higher confidence during AI-assisted decisions was tied to increased AI reliance. These findings highlight the dual nature of AI integration in clinical workflows: improving performance while introducing risks of bias-driven diagnostic errors.

Christof A. Bertram, Viktoria Weiss, Jonas Ammeling et al.

Supervised deep learning (DL) receives great interest for automated analysis of microscopic images with an increasing body of literature supporting its potential. The development and validation of those DL models relies heavily on the availability of high-quality, large-scale datasets. However, creating such datasets is a complex and resource-intensive process, often hindered by challenges such as time constraints, domain variability, and risks of bias in image collection and label creation. This review provides a comprehensive guide to the critical steps in dataset creation, including: 1) image acquisition, 2) selection of annotation software, and 3) annotation creation. In addition to ensuring a sufficiently large number of images, it is crucial to address sources of image variability (domain shifts) - such as those related to slide preparation and digitization - that could lead to algorithmic errors if not adequately represented in the training data. Key quality criteria for annotations are the three "C"s: correctness, completeness, and consistency. This review explores methods to enhance annotation quality through the use of advanced techniques that mitigate the limitations of single annotators. To support dataset creators, a standard operating procedure (SOP) is provided as supplemental material, outlining best practices for dataset development. Furthermore, the article underscores the importance of open datasets in driving innovation and enhancing reproducibility of DL research. By addressing the challenges and offering practical recommendations, this review aims to advance the creation of and availability to high-quality, large-scale datasets, ultimately contributing to the development of generalizable and robust DL models for pathology applications.

Frauke Wilm, Mathias Öttl, Marc Aubreville et al.

Recent advances in computer-aided diagnosis for histopathology have been largely driven by the use of deep learning models for automated image analysis. While these networks can perform on par with medical experts, their performance can be impeded by out-of-distribution data. The Cross-Organ and Cross-Scanner Adenocarcinoma Segmentation (COSAS) challenge aimed to address the task of cross-domain adenocarcinoma segmentation in the presence of morphological and scanner-induced domain shifts. In this paper, we present a U-Net-based segmentation framework designed to tackle this challenge. Our approach achieved segmentation scores of 0.8020 for the cross-organ track and 0.8527 for the cross-scanner track on the final challenge test sets, ranking it the best-performing submission.

Jonas Ammeling, Moritz Hecker, Jonathan Ganz et al.

The volume-corrected mitotic index (M/V-Index) was shown to provide prognostic value in invasive breast carcinomas. However, despite its prognostic significance, it is not established as the standard method for assessing aggressive biological behaviour, due to the high additional workload associated with determining the epithelial proportion. In this work, we show that using a deep learning pipeline solely trained with an annotation-free, immunohistochemistry-based approach, provides accurate estimations of epithelial segmentation in canine breast carcinomas. We compare our automatic framework with the manually annotated M/V-Index in a study with three board-certified pathologists. Our results indicate that the deep learning-based pipeline shows expert-level performance, while providing time efficiency and reproducibility.

Dominic Rampas, Pablo Pernias, Marc Aubreville

Recent advancements in the domain of text-to-image synthesis have culminated in a multitude of enhancements pertaining to quality, fidelity, and diversity. Contemporary techniques enable the generation of highly intricate visuals which rapidly approach near-photorealistic quality. Nevertheless, as progress is achieved, the complexity of these methodologies increases, consequently intensifying the comprehension barrier between individuals within the field and those external to it. In an endeavor to mitigate this disparity, we propose a streamlined approach for text-to-image generation, which encompasses both the training paradigm and the sampling process. Despite its remarkable simplicity, our method yields aesthetically pleasing images with few sampling iterations, allows for intriguing ways for conditioning the model, and imparts advantages absent in state-of-the-art techniques. To demonstrate the efficacy of this approach in achieving outcomes comparable to existing works, we have trained a one-billion parameter text-conditional model, which we refer to as "Paella". In the interest of fostering future exploration in this field, we have made our source code and models publicly accessible for the research community.

Sweta Banerjee, Viktoria Weiss, Taryn A. Donovan et al.

Atypical mitosis marks a deviation in the cell division process that has been shown be an independent prognostic marker for tumor malignancy. However, atypical mitosis classification remains challenging due to low prevalence, at times subtle morphological differences from normal mitotic figures, low inter-rater agreement among pathologists, and class imbalance in datasets. Building on the Atypical Mitosis dataset for Breast Cancer (AMi-Br), this study presents a comprehensive benchmark comparing deep learning approaches for automated atypical mitotic figure (AMF) classification, including end-to-end trained deep learning models, foundation models with linear probing, and foundation models fine-tuned with low-rank adaptation (LoRA). For rigorous evaluation, we further introduce two new held-out AMF datasets - AtNorM-Br, a dataset of mitotic figures from the TCGA breast cancer cohort, and AtNorM-MD, a multi-domain dataset of mitotic figures from a subset of the MIDOG++ training set. We found average balanced accuracy values of up to 0.8135, 0.7788, and 0.7723 on the in-domain AMi-Br and the out-of-domain AtNorm-Br and AtNorM-MD datasets, respectively. Our work shows that atypical mitotic figure classification, while being a challenging problem, can be effectively addressed through the use of recent advances in transfer learning and model fine-tuning techniques. We make all code and data used in this paper available in this github repository: https://github.com/DeepMicroscopy/AMi-Br_Benchmark.

Jingna Qiu, Nishanth Jain, Jonas Ammeling et al.

Recent advances in Vision-Language Models (VLMs) in histopathology, such as CONCH and QuiltNet, have demonstrated impressive zero-shot classification capabilities across various tasks. However, their general-purpose design may lead to suboptimal performance in specific downstream applications. While supervised fine-tuning methods address this issue, they require manually labeled samples for adaptation. This paper investigates annotation-free adaptation of VLMs through continued pretraining on domain- and task-relevant image-caption pairs extracted from existing databases. Our experiments on two VLMs, CONCH and QuiltNet, across three downstream tasks reveal that these pairs substantially enhance both zero-shot and few-shot performance. Notably, with larger training sizes, continued pretraining matches the performance of few-shot methods while eliminating manual labeling. Its effectiveness, task-agnostic design, and annotation-free workflow make it a promising pathway for adapting VLMs to new histopathology tasks. Code is available at https://github.com/DeepMicroscopy/Annotation-free-VLM-specialization.

Christof A. Bertram, Taryn A. Donovan, Marco Tecilla et al.

Tumor cells with two nuclei (binucleated cells, BiNC) or more nuclei (multinucleated cells, MuNC) indicate an increased amount of cellular genetic material which is thought to facilitate oncogenesis, tumor progression and treatment resistance. In canine cutaneous mast cell tumors (ccMCT), binucleation and multinucleation are parameters used in cytologic and histologic grading schemes (respectively) which correlate with poor patient outcome. For this study, we created the first open source data-set with 19,983 annotations of BiNC and 1,416 annotations of MuNC in 32 histological whole slide images of ccMCT. Labels were created by a pathologist and an algorithmic-aided labeling approach with expert review of each generated candidate. A state-of-the-art deep learning-based model yielded an $F_1$ score of 0.675 for BiNC and 0.623 for MuNC on 11 test whole slide images. In regions of interest ($2.37 mm^2$) extracted from these test images, 6 pathologists had an object detection performance between 0.270 - 0.526 for BiNC and 0.316 - 0.622 for MuNC, while our model archived an $F_1$ score of 0.667 for BiNC and 0.685 for MuNC. This open dataset can facilitate development of automated image analysis for this task and may thereby help to promote standardization of this facet of histologic tumor prognostication.

Christian Marzahl, Marc Aubreville, Christof A. Bertram et al.

In many research areas, scientific progress is accelerated by multidisciplinary access to image data and their interdisciplinary annotation. However, keeping track of these annotations to ensure a high-quality multi-purpose data set is a challenging and labour intensive task. We developed the open-source online platform EXACT (EXpert Algorithm Collaboration Tool) that enables the collaborative interdisciplinary analysis of images from different domains online and offline. EXACT supports multi-gigapixel medical whole slide images as well as image series with thousands of images. The software utilises a flexible plugin system that can be adapted to diverse applications such as counting mitotic figures with a screening mode, finding false annotations on a novel validation view, or using the latest deep learning image analysis technologies. This is combined with a version control system which makes it possible to keep track of changes in the data sets and, for example, to link the results of deep learning experiments to specific data set versions. EXACT is freely available and has already been successfully applied to a broad range of annotation tasks, including highly diverse applications like deep learning supported cytology scoring, interdisciplinary multi-centre whole slide image tumour annotation, and highly specialised whale sound spectroscopy clustering.

Marc Aubreville, Christof Bertram, Robert Klopfleisch et al.

Large-scale image data such as digital whole-slide histology images pose a challenging task at annotation software solutions. Today, a number of good solutions with varying scopes exist. For cell annotation, however, we find that many do not match the prerequisites for fast annotations. Especially in the field of mitosis detection, it is assumed that detection accuracy could significantly benefit from larger annotation databases that are currently however very troublesome to produce. Further, multiple independent (blind) expert labels are a big asset for such databases, yet there is currently no tool for this kind of annotation available. To ease this tedious process of expert annotation and grading, we introduce SlideRunner, an open source annotation and visualization tool for digital histopathology, developed in close cooperation with two pathologists. SlideRunner is capable of setting annotations like object centers (for e.g. cells) as well as object boundaries (e.g. for tumor outlines). It provides single-click annotations as well as a blind mode for multi-annotations, where the expert is directly shown the microscopy image containing the cells that he has not yet rated.

Nils Porsche, Flurin Müller-Diesing, Sweta Banerjee et al.

Confocal laser endomicroscopy (CLE) is a non-invasive, real-time imaging modality that can be used for in-situ, in-vivo imaging and the microstructural analysis of mucous structures. The diagnosis using CLE is, however, complicated by images being hard to interpret for non-experienced physicians. Utilizing machine learning as an augmentative tool would hence be beneficial, but is complicated by the shortage of histopathology-correlated CLE imaging sequences with respect to the plurality of patterns in this domain, leading to overfitting of machine learning models. To overcome this, self-supervised learning (SSL) can be employed on larger unlabeled datasets. CLE is a video-based modality with high inter-frame correlation, leading to a non-stratified data distribution for SSL training. In this work, we propose a filter functionality on CLE video sequences to reduce the dataset redundancy in SSL training and improve SSL training convergence and training efficiency. We use four state-of-the-art baseline networks and a SSL teacher-student network with a vision transformer small backbone for the evaluation. These networks were evaluated on downstream tasks for a sinonasal tumor dataset and a squamous cell carcinoma of the skin dataset. On both datasets, we found the highest test accuracy on the filtered SSL-pretrained model, with 67.48% and 73.52%, both considerably outperforming their non-SSL baselines. Our results show that SSL is an effective method for CLE pretraining. Further, we show that our proposed CLE video filter can be utilized to improve training efficiency in self-supervised scenarios, resulting in a reduction of 67% in training time.

Christof A. Bertram, Viktoria Weiss, Taryn A. Donovan et al.

Assessment of the density of mitotic figures (MFs) in histologic tumor sections is an important prognostic marker for many tumor types, including breast cancer. Recently, it has been reported in multiple works that the quantity of MFs with an atypical morphology (atypical MFs, AMFs) might be an independent prognostic criterion for breast cancer. AMFs are an indicator of mutations in the genes regulating the cell cycle and can lead to aberrant chromosome constitution (aneuploidy) of the tumor cells. To facilitate further research on this topic using pattern recognition, we present the first ever publicly available dataset of atypical and normal MFs (AMi-Br). For this, we utilized two of the most popular MF datasets (MIDOG 2021 and TUPAC) and subclassified all MFs using a three expert majority vote. Our final dataset consists of 3,720 MFs, split into 832 AMFs (22.4%) and 2,888 normal MFs (77.6%) across all 223 tumor cases in the combined set. We provide baseline classification experiments to investigate the consistency of the dataset, using a Monte Carlo cross-validation and different strategies to combat class imbalance. We found an averaged balanced accuracy of up to 0.806 when using a patch-level data set split, and up to 0.713 when using a patient-level split.

Marc Aubreville, Jonathan Ganz, Jonas Ammeling et al.

The QUILT-1M dataset is the first openly available dataset containing images harvested from various online sources. While it provides a huge data variety, the image quality and composition is highly heterogeneous, impacting its utility for text-conditional image synthesis. We propose an automatic pipeline that provides predictions of the most common impurities within the images, e.g., visibility of narrators, desktop environment and pathology software, or text within the image. Additionally, we propose to use semantic alignment filtering of the image-text pairs. Our findings demonstrate that by rigorously filtering the dataset, there is a substantial enhancement of image fidelity in text-to-image tasks.

Frauke Wilm, Jonas Ammeling, Mathias Öttl et al.

The U-net architecture has significantly impacted deep learning-based segmentation of medical images. Through the integration of long-range skip connections, it facilitated the preservation of high-resolution features. Out-of-distribution data can, however, substantially impede the performance of neural networks. Previous works showed that the trained network layers differ in their susceptibility to this domain shift, e.g., shallow layers are more affected than deeper layers. In this work, we investigate the implications of this observation of layer sensitivity to domain shifts of U-net-style segmentation networks. By copying features of shallow layers to corresponding decoder blocks, these bear the risk of re-introducing domain-specific information. We used a synthetic dataset to model different levels of data distribution shifts and evaluated the impact on downstream segmentation performance. We quantified the inherent domain susceptibility of each network layer, using the Hellinger distance. These experiments confirmed the higher domain susceptibility of earlier network layers. When gradually removing skip connections, a decrease in domain susceptibility of deeper layers could be observed. For downstream segmentation performance, the original U-net outperformed the variant without any skip connections. The best performance, however, was achieved when removing the uppermost skip connection - not only in the presence of domain shifts but also for in-domain test data. We validated our results on three clinical datasets - two histopathology datasets and one magnetic resonance dataset - with performance increases of up to 10% in-domain and 13% cross-domain when removing the uppermost skip connection.

Chloé Puget, Jonathan Ganz, Julian Ostermaier et al.

Numerous prognostic factors are currently assessed histopathologically in biopsies of canine mast cell tumors to evaluate clinical behavior. In addition, PCR analysis of the c-Kit exon 11 mutational status is often performed to evaluate the potential success of a tyrosine kinase inhibitor therapy. This project aimed at training deep learning models (DLMs) to identify the c-Kit-11 mutational status of MCTs solely based on morphology without additional molecular analysis. HE slides of 195 mutated and 173 non-mutated tumors were stained consecutively in two different laboratories and scanned with three different slide scanners. This resulted in six different datasets (stain-scanner variations) of whole slide images. DLMs were trained with single and mixed datasets and their performances was assessed under scanner and staining domain shifts. The DLMs correctly classified HE slides according to their c-Kit 11 mutation status in, on average, 87% of cases for the best-suited stain-scanner variant. A relevant performance drop could be observed when the stain-scanner combination of the training and test dataset differed. Multi-variant datasets improved the average accuracy but did not reach the maximum accuracy of algorithms trained and tested on the same stain-scanner variant. In summary, DLM-assisted morphological examination of MCTs can predict c-Kit-exon 11 mutational status of MCTs with high accuracy. However, the recognition performance is impeded by a change of scanner or staining protocol. Larger data sets with higher numbers of scans originating from different laboratories and scanners may lead to more robust DLMs to identify c-Kit mutations in HE slides.

Jonas Ammeling, Jonathan Ganz, Emely Rosbach et al.

The performance of deep learning models is known to scale with data quantity and diversity. In pathology, as in many other medical imaging domains, the availability of labeled images for a specific task is often limited. Self-supervised learning techniques have enabled the use of vast amounts of unlabeled data to train large-scale neural networks, i.e., foundation models, that can address the limited data problem by providing semantically rich feature vectors that can generalize well to new tasks with minimal training effort increasing model performance and robustness. In this work, we investigate the use of foundation models for mitotic figure classification. The mitotic count, which can be derived from this classification task, is an independent prognostic marker for specific tumors and part of certain tumor grading systems. In particular, we investigate the data scaling laws on multiple current foundation models and evaluate their robustness to unseen tumor domains. Next to the commonly used linear probing paradigm, we also adapt the models using low-rank adaptation (LoRA) of their attention mechanisms. We compare all models against end-to-end-trained baselines, both CNNs and Vision Transformers. Our results demonstrate that LoRA-adapted foundation models provide superior performance to those adapted with standard linear probing, reaching performance levels close to 100% data availability with only 10% of training data. Furthermore, LoRA-adaptation of the most recent foundation models almost closes the out-of-domain performance gap when evaluated on unseen tumor domains. However, full fine-tuning of traditional architectures still yields competitive performance.

Jonathan Ganz, Jonas Ammeling, Emely Rosbach et al.

Foundation models (FMs), i.e., models trained on a vast amount of typically unlabeled data, have become popular and available recently for the domain of histopathology. The key idea is to extract semantically rich vectors from any input patch, allowing for the use of simple subsequent classification networks potentially reducing the required amounts of labeled data, and increasing domain robustness. In this work, we investigate to which degree this also holds for mitotic figure classification. Utilizing two popular public mitotic figure datasets, we compared linear probing of five publicly available FMs against models trained on ImageNet and a simple ResNet50 end-to-end-trained baseline. We found that the end-to-end-trained baseline outperformed all FM-based classifiers, regardless of the amount of data provided. Additionally, we did not observe the FM-based classifiers to be more robust against domain shifts, rendering both of the above assumptions incorrect.

Jonathan Ganz, Christian Marzahl, Jonas Ammeling et al.

The count of mitotic figures (MFs) observed in hematoxylin and eosin (H&E)-stained slides is an important prognostic marker as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. Deep learning algorithms can standardize this task, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithm's performance. Unlike H&E, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H&E stain alone, the use of this ground truth could potentially introduce noise into the H&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. We found that the annotators' object-level agreement increased when using PHH3-assisted labeling. Subsequently, MF detectors were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. Additionally, a novel dual-stain MF detector was developed to investigate the interpretation-shift of PHH3-assisted labels used in H&E, which clearly outperformed single-stain detectors. However, the PHH3-assisted labels did not have a positive effect on solely H&E-based models. The high performance of our dual-input detector reveals an information mismatch between the H&E and PHH3-stained images as the cause of this effect.

Jonathan Ganz, Jonas Ammeling, Samir Jabari et al.

In numerous studies, deep learning algorithms have proven their potential for the analysis of histopathology images, for example, for revealing the subtypes of tumors or the primary origin of metastases. These models require large datasets for training, which must be anonymized to prevent possible patient identity leaks. This study demonstrates that even relatively simple deep learning algorithms can re-identify patients in large histopathology datasets with substantial accuracy. We evaluated our algorithms on two TCIA datasets including lung squamous cell carcinoma (LSCC) and lung adenocarcinoma (LUAD). We also demonstrate the algorithm's performance on an in-house dataset of meningioma tissue. We predicted the source patient of a slide with F1 scores of 50.16 % and 52.30 % on the LSCC and LUAD datasets, respectively, and with 62.31 % on our meningioma dataset. Based on our findings, we formulated a risk assessment scheme to estimate the risk to the patient's privacy prior to publication.

Frauke Wilm, Marco Fragoso, Christian Marzahl et al.

Due to morphological similarities, the differentiation of histologic sections of cutaneous tumors into individual subtypes can be challenging. Recently, deep learning-based approaches have proven their potential for supporting pathologists in this regard. However, many of these supervised algorithms require a large amount of annotated data for robust development. We present a publicly available dataset of 350 whole slide images of seven different canine cutaneous tumors complemented by 12,424 polygon annotations for 13 histologic classes, including seven cutaneous tumor subtypes. In inter-rater experiments, we show a high consistency of the provided labels, especially for tumor annotations. We further validate the dataset by training a deep neural network for the task of tissue segmentation and tumor subtype classification. We achieve a class-averaged Jaccard coefficient of 0.7047, and 0.9044 for tumor in particular. For classification, we achieve a slide-level accuracy of 0.9857. Since canine cutaneous tumors possess various histologic homologies to human tumors the added value of this dataset is not limited to veterinary pathology but extends to more general fields of application.

Frauke Wilm, Christian Marzahl, Katharina Breininger et al.

Assessing the Mitotic Count has a known high degree of intra- and inter-rater variability. Computer-aided systems have proven to decrease this variability and reduce labeling time. These systems, however, are generally highly dependent on their training domain and show poor applicability to unseen domains. In histopathology, these domain shifts can result from various sources, including different slide scanning systems used to digitize histologic samples. The MItosis DOmain Generalization challenge focused on this specific domain shift for the task of mitotic figure detection. This work presents a mitotic figure detection algorithm developed as a baseline for the challenge, based on domain adversarial training. On the challenge's test set, the algorithm scored an F$_1$ score of 0.7183. The corresponding network weights and code for implementing the network are made publicly available.

Christian Marzahl, Jenny Hill, Jason Stayt et al.

Pulmonary hemorrhage (P-Hem) occurs among multiple species and can have various causes. Cytology of bronchoalveolarlavage fluid (BALF) using a 5-tier scoring system of alveolar macrophages based on their hemosiderin content is considered the most sensitive diagnostic method. We introduce a novel, fully annotated multi-species P-Hem dataset which consists of 74 cytology whole slide images (WSIs) with equine, feline and human samples. To create this high-quality and high-quantity dataset, we developed an annotation pipeline combining human expertise with deep learning and data visualisation techniques. We applied a deep learning-based object detection approach trained on 17 expertly annotated equine WSIs, to the remaining 39 equine, 12 human and 7 feline WSIs. The resulting annotations were semi-automatically screened for errors on multiple types of specialised annotation maps and finally reviewed by a trained pathologists. Our dataset contains a total of 297,383 hemosiderophages classified into five grades. It is one of the largest publicly availableWSIs datasets with respect to the number of annotations, the scanned area and the number of species covered.

Jonathan Ganz, Tobias Kirsch, Lucas Hoffmann et al.

Meningioma is one of the most prevalent brain tumors in adults. To determine its malignancy, it is graded by a pathologist into three grades according to WHO standards. This grade plays a decisive role in treatment, and yet may be subject to inter-rater discordance. In this work, we present and compare three approaches towards fully automatic meningioma grading from histology whole slide images. All approaches are following a two-stage paradigm, where we first identify a region of interest based on the detection of mitotic figures in the slide using a state-of-the-art object detection deep learning network. This region of highest mitotic rate is considered characteristic for biological tumor behavior. In the second stage, we calculate a score corresponding to tumor malignancy based on information contained in this region using three different settings. In a first approach, image patches are sampled from this region and regression is based on morphological features encoded by a ResNet-based network. We compare this to learning a logistic regression from the determined mitotic count, an approach which is easily traceable and explainable. Lastly, we combine both approaches in a single network. We trained the pipeline on 951 slides from 341 patients and evaluated them on a separate set of 141 slides from 43 patients. All approaches yield a high correlation to the WHO grade. The logistic regression and the combined approach had the best results in our experiments, yielding correct predictions in 32 and 33 of all cases, respectively, with the image-based approach only predicting 25 cases correctly. Spearman's correlation was 0.716, 0.792 and 0.790 respectively. It may seem counterintuitive at first that morphological features provided by image patches do not improve model performance. Yet, this mirrors the criteria of the grading scheme, where mitotic count is the only unequivocal parameter.

Marc Aubreville, Christof Bertram, Mitko Veta et al.

Automated detection of mitotic figures in histopathology images has seen vast improvements, thanks to modern deep learning-based pipelines. Application of these methods, however, is in practice limited by strong variability of images between labs. This results in a domain shift of the images, which causes a performance drop of the models. Hypothesizing that the scanner device plays a decisive role in this effect, we evaluated the susceptibility of a standard mitosis detection approach to the domain shift introduced by using a different whole slide scanner. Our work is based on the MICCAI-MIDOG challenge 2021 data set, which includes 200 tumor cases of human breast cancer and four scanners. Our work indicates that the domain shift induced not by biochemical variability but purely by the choice of acquisition device is underestimated so far. Models trained on images of the same scanner yielded an average F1 score of 0.683, while models trained on a single other scanner only yielded an average F1 score of 0.325. Training on another multi-domain mitosis dataset led to mean F1 scores of 0.52. We found this not to be reflected by domain-shifts measured as proxy A distance-derived metric.

Christian Marzahl, Christof A. Bertram, Frauke Wilm et al.

Asthma is a chronic inflammatory disorder of the lower respiratory tract and naturally occurs in humans and animals including horses. The annotation of an asthma microscopy whole slide image (WSI) is an extremely labour-intensive task due to the hundreds of thousands of cells per WSI. To overcome the limitation of annotating WSI incompletely, we developed a training pipeline which can train a deep learning-based object detection model with partially annotated WSIs and compensate class imbalances on the fly. With this approach we can freely sample from annotated WSIs areas and are not restricted to fully annotated extracted sub-images of the WSI as with classical approaches. We evaluated our pipeline in a cross-validation setup with a fixed training set using a dataset of six equine WSIs of which four are partially annotated and used for training, and two fully annotated WSI are used for validation and testing. Our WSI-based training approach outperformed classical sub-image-based training methods by up to 15\% $mAP$ and yielded human-like performance when compared to the annotations of ten trained pathologists.

Frauke Wilm, Christof A. Bertram, Christian Marzahl et al.

Density of mitotic figures in histologic sections is a prognostically relevant characteristic for many tumours. Due to high inter-pathologist variability, deep learning-based algorithms are a promising solution to improve tumour prognostication. Pathologists are the gold standard for database development, however, labelling errors may hamper development of accurate algorithms. In the present work we evaluated the benefit of multi-expert consensus (n = 3, 5, 7, 9, 11) on algorithmic performance. While training with individual databases resulted in highly variable F$_1$ scores, performance was notably increased and more consistent when using the consensus of three annotators. Adding more annotators only resulted in minor improvements. We conclude that databases by few pathologists and high label accuracy may be the best compromise between high algorithmic performance and time investment.

Marc Aubreville, Christof A. Bertram, Taryn A. Donovan et al.

Canine mammary carcinoma (CMC) has been used as a model to investigate the pathogenesis of human breast cancer and the same grading scheme is commonly used to assess tumor malignancy in both. One key component of this grading scheme is the density of mitotic figures (MF). Current publicly available datasets on human breast cancer only provide annotations for small subsets of whole slide images (WSIs). We present a novel dataset of 21 WSIs of CMC completely annotated for MF. For this, a pathologist screened all WSIs for potential MF and structures with a similar appearance. A second expert blindly assigned labels, and for non-matching labels, a third expert assigned the final labels. Additionally, we used machine learning to identify previously undetected MF. Finally, we performed representation learning and two-dimensional projection to further increase the consistency of the annotations. Our dataset consists of 13,907 MF and 36,379 hard negatives. We achieved a mean F1-score of 0.791 on the test set and of up to 0.696 on a human breast cancer dataset.

Christof A. Bertram, Mitko Veta, Christian Marzahl et al.

Pathologist-defined labels are the gold standard for histopathological data sets, regardless of well-known limitations in consistency for some tasks. To date, some datasets on mitotic figures are available and were used for development of promising deep learning-based algorithms. In order to assess robustness of those algorithms and reproducibility of their methods it is necessary to test on several independent datasets. The influence of different labeling methods of these available datasets is currently unknown. To tackle this, we present an alternative set of labels for the images of the auxiliary mitosis dataset of the TUPAC16 challenge. Additional to manual mitotic figure screening, we used a novel, algorithm-aided labeling process, that allowed to minimize the risk of missing rare mitotic figures in the images. All potential mitotic figures were independently assessed by two pathologists. The novel, publicly available set of labels contains 1,999 mitotic figures (+28.80%) and additionally includes 10,483 labels of cells with high similarities to mitotic figures (hard examples). We found significant difference comparing F_1 scores between the original label set (0.549) and the new alternative label set (0.735) using a standard deep learning object detection architecture. The models trained on the alternative set showed higher overall confidence values, suggesting a higher overall label consistency. Findings of the present study show that pathologists-defined labels may vary significantly resulting in notable difference in the model performance. Comparison of deep learning-based algorithms between independent datasets with different labeling methods should be done with caution.

Christian Marzahl, Christof A. Bertram, Marc Aubreville et al.

Deep-learning-based pipelines have shown the potential to revolutionalize microscopy image diagnostics by providing visual augmentations to a trained pathology expert. However, to match human performance, the methods rely on the availability of vast amounts of high-quality labeled data, which poses a significant challenge. To circumvent this, augmented labeling methods, also known as expert-algorithm-collaboration, have recently become popular. However, potential biases introduced by this operation mode and their effects for training neuronal networks are not entirely understood. This work aims to shed light on some of the effects by providing a case study for three pathologically relevant diagnostic settings. Ten trained pathology experts performed a labeling tasks first without and later with computer-generated augmentation. To investigate different biasing effects, we intentionally introduced errors to the augmentation. Furthermore, we developed a novel loss function which incorporates the experts' annotation consensus in the training of a deep learning classifier. In total, the pathology experts annotated 26,015 cells on 1,200 images in this novel annotation study. Backed by this extensive data set, we found that the consensus of multiple experts and the deep learning classifier accuracy, was significantly increased in the computer-aided setting, versus the unaided annotation. However, a significant percentage of the deliberately introduced false labels was not identified by the experts. Additionally, we showed that our loss function profited from multiple experts and outperformed conventional loss functions. At the same time, systematic errors did not lead to a deterioration of the trained classifier accuracy. Furthermore, a classifier trained with annotations from a single expert with computer-aided support can outperform the combined annotations from up to nine experts.

Hendrik Schröter, Tobias Rosenkranz, Alberto N. Escalante B. et al.

Noise reduction is an important part of modern hearing aids and is included in most commercially available devices. Deep learning-based state-of-the-art algorithms, however, either do not consider real-time and frequency resolution constrains or result in poor quality under very noisy conditions. To improve monaural speech enhancement in noisy environments, we propose CLCNet, a framework based on complex valued linear coding. First, we define complex linear coding (CLC) motivated by linear predictive coding (LPC) that is applied in the complex frequency domain. Second, we propose a framework that incorporates complex spectrogram input and coefficient output. Third, we define a parametric normalization for complex valued spectrograms that complies with low-latency and on-line processing. Our CLCNet was evaluated on a mixture of the EUROM database and a real-world noise dataset recorded with hearing aids and compared to traditional real-valued Wiener-Filter gains.

Christian Marzahl, Marc Aubreville, Christof A. Bertram et al.

Modern, state-of-the-art deep learning approaches yield human like performance in numerous object detection and classification tasks. The foundation for their success is the availability of training datasets of substantially high quantity, which are expensive to create, especially in the field of medical imaging. Recently, crowdsourcing has been applied to create large datasets for a broad range of disciplines. This study aims to explore the challenges and opportunities of crowd-algorithm collaboration for the object detection task of grading cytology whole slide images. We compared the classical crowdsourcing performance of twenty participants with their results from crowd-algorithm collaboration. All participants performed both modes in random order on the same twenty images. Additionally, we introduced artificial systematic flaws into the precomputed annotations to estimate a bias towards accepting precomputed annotations. We gathered 9524 annotations on 800 images from twenty participants organised into four groups in concordance to their level of expertise with cytology. The crowd-algorithm mode improved on average the participants' classification accuracy by 7%, the mean average precision by 8% and the inter-observer Fleiss' kappa score by 20%, and reduced the time spent by 31%. However, two thirds of the artificially modified false labels were not recognised as such by the contributors. This study shows that crowd-algorithm collaboration is a promising new approach to generate large datasets when it is ensured that a carefully designed setup eliminates potential biases.

Marc Aubreville, Christof A. Bertram, Samir Jabari et al.

For histopathological tumor assessment, the count of mitotic figures per area is an important part of prognostication. Algorithmic approaches - such as for mitotic figure identification - have significantly improved in recent times, potentially allowing for computer-augmented or fully automatic screening systems in the future. This trend is further supported by whole slide scanning microscopes becoming available in many pathology labs and could soon become a standard imaging tool. For an application in broader fields of such algorithms, the availability of mitotic figure data sets of sufficient size for the respective tissue type and species is an important precondition, that is, however, rarely met. While algorithmic performance climbed steadily for e.g. human mammary carcinoma, thanks to several challenges held in the field, for most tumor types, data sets are not available. In this work, we assess domain transfer of mitotic figure recognition using domain adversarial training on four data sets, two from dogs and two from humans. We were able to show that domain adversarial training considerably improves accuracy when applying mitotic figure classification learned from the canine on the human data sets (up to +12.8% in accuracy) and is thus a helpful method to transfer knowledge from existing data sets to new tissue types and species.

Christian Marzahl, Marc Aubreville, Christof A. Bertram et al.

Purpose: Exercise-induced pulmonary hemorrhage (EIPH) is a common syndrome in sport horses with negative impact on performance. Cytology of bronchoalveolar lavage fluid by use of a scoring system is considered the most sensitive diagnostic method. Macrophages are classified depending on the degree of cytoplasmic hemosiderin content. The current gold standard is manual grading, which is however monotonous and time-consuming. Methods: We evaluated state-of-the-art deep learning-based methods for single cell macrophage classification and compared them against the performance of nine cytology experts and evaluated inter- and intra-observer variability. Additionally, we evaluated object detection methods on a novel data set of 17 completely annotated cytology whole slide images (WSI) containing 78,047 hemosiderophages. Resultsf: Our deep learning-based approach reached a concordance of 0.85, partially exceeding human expert concordance (0.68 to 0.86, $μ$=0.73, $σ$ =0.04). Intra-observer variability was high (0.68 to 0.88) and inter-observer concordance was moderate (Fleiss kappa = 0.67). Our object detection approach has a mean average precision of 0.66 over the five classes from the whole slide gigapixel image and a computation time of below two minutes. Conclusion: To mitigate the high inter- and intra-rater variability, we propose our automated object detection pipeline, enabling accurate, reproducible and quick EIPH scoring in WSI.