Jens Petersen

Lena Maier-Hein, Annika Reinke, Patrick Godau et al. · utoronto

Increasing evidence shows that flaws in machine learning (ML) algorithm validation are an underestimated global problem. Particularly in automatic biomedical image analysis, chosen performance metrics often do not reflect the domain interest, thus failing to adequately measure scientific progress and hindering translation of ML techniques into practice. To overcome this, our large international expert consortium created Metrics Reloaded, a comprehensive framework guiding researchers in the problem-aware selection of metrics. Following the convergence of ML methodology across application domains, Metrics Reloaded fosters the convergence of validation methodology. The framework was developed in a multi-stage Delphi process and is based on the novel concept of a problem fingerprint - a structured representation of the given problem that captures all aspects that are relevant for metric selection, from the domain interest to the properties of the target structure(s), data set and algorithm output. Based on the problem fingerprint, users are guided through the process of choosing and applying appropriate validation metrics while being made aware of potential pitfalls. Metrics Reloaded targets image analysis problems that can be interpreted as a classification task at image, object or pixel level, namely image-level classification, object detection, semantic segmentation, and instance segmentation tasks. To improve the user experience, we implemented the framework in the Metrics Reloaded online tool, which also provides a point of access to explore weaknesses, strengths and specific recommendations for the most common validation metrics. The broad applicability of our framework across domains is demonstrated by an instantiation for various biological and medical image analysis use cases.

Saikat Roy, Gregor Koehler, Constantin Ulrich et al.

There has been exploding interest in embracing Transformer-based architectures for medical image segmentation. However, the lack of large-scale annotated medical datasets make achieving performances equivalent to those in natural images challenging. Convolutional networks, in contrast, have higher inductive biases and consequently, are easily trainable to high performance. Recently, the ConvNeXt architecture attempted to modernize the standard ConvNet by mirroring Transformer blocks. In this work, we improve upon this to design a modernized and scalable convolutional architecture customized to challenges of data-scarce medical settings. We introduce MedNeXt, a Transformer-inspired large kernel segmentation network which introduces - 1) A fully ConvNeXt 3D Encoder-Decoder Network for medical image segmentation, 2) Residual ConvNeXt up and downsampling blocks to preserve semantic richness across scales, 3) A novel technique to iteratively increase kernel sizes by upsampling small kernel networks, to prevent performance saturation on limited medical data, 4) Compound scaling at multiple levels (depth, width, kernel size) of MedNeXt. This leads to state-of-the-art performance on 4 tasks on CT and MRI modalities and varying dataset sizes, representing a modernized deep architecture for medical image segmentation. Our code is made publicly available at: https://github.com/MIC-DKFZ/MedNeXt.

Annika Reinke, Minu D. Tizabi, Michael Baumgartner et al.

Validation metrics are key for the reliable tracking of scientific progress and for bridging the current chasm between artificial intelligence (AI) research and its translation into practice. However, increasing evidence shows that particularly in image analysis, metrics are often chosen inadequately in relation to the underlying research problem. This could be attributed to a lack of accessibility of metric-related knowledge: While taking into account the individual strengths, weaknesses, and limitations of validation metrics is a critical prerequisite to making educated choices, the relevant knowledge is currently scattered and poorly accessible to individual researchers. Based on a multi-stage Delphi process conducted by a multidisciplinary expert consortium as well as extensive community feedback, the present work provides the first reliable and comprehensive common point of access to information on pitfalls related to validation metrics in image analysis. Focusing on biomedical image analysis but with the potential of transfer to other fields, the addressed pitfalls generalize across application domains and are categorized according to a newly created, domain-agnostic taxonomy. To facilitate comprehension, illustrations and specific examples accompany each pitfall. As a structured body of information accessible to researchers of all levels of expertise, this work enhances global comprehension of a key topic in image analysis validation.

Julian Schön, Raghavendra Selvan, Lotte Nygård et al.

Medical imaging plays a vital role in modern diagnostics and treatment. The temporal nature of disease or treatment progression often results in longitudinal data. Due to the cost and potential harm, acquiring large medical datasets necessary for deep learning can be difficult. Medical image synthesis could help mitigate this problem. However, until now, the availability of GANs capable of synthesizing longitudinal volumetric data has been limited. To address this, we use the recent advances in latent space-based image editing to propose a novel joint learning scheme to explicitly embed temporal dependencies in the latent space of GANs. This, in contrast to previous methods, allows us to synthesize continuous, smooth, and high-quality longitudinal volumetric data with limited supervision. We show the effectiveness of our approach on three datasets containing different longitudinal dependencies. Namely, modeling a simple image transformation, breathing motion, and tumor regression, all while showing minimal disentanglement. The implementation is made available online at https://github.com/julschoen/Temp-GAN.

Abraham George Smith, Denis Kutnár, Ivan Richter Vogelius et al.

Increased organ at risk segmentation accuracy is required to reduce cost and complications for patients receiving radiotherapy treatment. Some deep learning methods for the segmentation of organs at risk use a two stage process where a localisation network first crops an image to the relevant region and then a locally specialised network segments the cropped organ of interest. We investigate the accuracy improvements brought about by such a localisation stage by comparing to a single-stage baseline network trained on full resolution images. We find that localisation approaches can improve both training time and stability and a two stage process involving both a localisation and organ segmentation network provides a significant increase in segmentation accuracy for the spleen, pancreas and heart from the Medical Segmentation Decathlon dataset. We also observe increased benefits of localisation for smaller organs. Source code that recreates the main results is available at \href{https://github.com/Abe404/localise_to_segment}{this https URL}.

Carsten T. Lüth, David Zimmerer, Gregor Koehler et al.

Unsupervised anomaly detection in medical imaging aims to detect and localize arbitrary anomalies without requiring annotated anomalous data during training. Often, this is achieved by learning a data distribution of normal samples and detecting anomalies as regions in the image which deviate from this distribution. Most current state-of-the-art methods use latent variable generative models operating directly on the images. However, generative models have been shown to mostly capture low-level features, s.a. pixel-intensities, instead of rich semantic features, which also applies to their representations. We circumvent this problem by proposing CRADL whose core idea is to model the distribution of normal samples directly in the low-dimensional representation space of an encoder trained with a contrastive pretext-task. By utilizing the representations of contrastive learning, we aim to fix the over-fixation on low-level features and learn more semantic-rich representations. Our experiments on anomaly detection and localization tasks using three distinct evaluation datasets show that 1) contrastive representations are superior to representations of generative latent variable models and 2) the CRADL framework shows competitive or superior performance to state-of-the-art.

Saikat Roy, Gregor Koehler, Michael Baumgartner et al.

Owing to success in the data-rich domain of natural images, Transformers have recently become popular in medical image segmentation. However, the pairing of Transformers with convolutional blocks in varying architectural permutations leaves their relative effectiveness to open interpretation. We introduce Transformer Ablations that replace the Transformer blocks with plain linear operators to quantify this effectiveness. With experiments on 8 models on 2 medical image segmentation tasks, we explore -- 1) the replaceable nature of Transformer-learnt representations, 2) Transformer capacity alone cannot prevent representational replaceability and works in tandem with effective design, 3) The mere existence of explicit feature hierarchies in transformer blocks is more beneficial than accompanying self-attention modules, 4) Major spatial downsampling before Transformer modules should be used with caution.

Julian Schön, Raghavendra Selvan, Jens Petersen

Generative models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) play an increasingly important role in medical image analysis. The latent spaces of these models often show semantically meaningful directions corresponding to human-interpretable image transformations. However, until now, their exploration for medical images has been limited due to the requirement of supervised data. Several methods for unsupervised discovery of interpretable directions in GAN latent spaces have shown interesting results on natural images. This work explores the potential of applying these techniques on medical images by training a GAN and a VAE on thoracic CT scans and using an unsupervised method to discover interpretable directions in the resulting latent space. We find several directions corresponding to non-trivial image transformations, such as rotation or breast size. Furthermore, the directions show that the generative models capture 3D structure despite being presented only with 2D data. The results show that unsupervised methods to discover interpretable directions in GANs generalize to VAEs and can be applied to medical images. This opens a wide array of future work using these methods in medical image analysis.

Malte Silbernagel, Albert Alonso, Jens Petersen et al.

Accurately predicting topologically correct masks remains a difficult task for general segmentation models, which often produce fragmented or disconnected outputs. Fixing these artifacts typically requires hand-crafted refinement rules or architectures specialized to a particular task. Here, we show that Neural Cellular Automata (NCA) can be directly re-purposed as an effective refinement mechanism, using local, iterative updates guided by image context to repair segmentation masks. By training on imperfect masks and ground truths, the automaton learns the structural properties of the target shape while relying solely on local information. When applied to coarse, globally predicted masks, the learned dynamics progressively reconnect broken regions, prune loose fragments and converge towards stable, topologically consistent results. We show how refinement NCA (rNCA) can be easily applied to repair common topological errors produced by different base segmentation models and tasks: for fragmented retinal vessels, it yields 2-3% gains in Dice/clDice and improves Betti errors, reducing $β_0$ errors by 60% and $β_1$ by 20%; for myocardium, it repairs 61.5% of broken cases in a zero-shot setting while lowering ASSD and HD by 19% and 16%, respectively. This showcases NCA as effective and broadly applicable refiners.

Denis Kutnár, Ivan R Vogelius, Katrin Elisabet Håkansson et al.

Locoregional recurrences (LRR) are still a frequent site of treatment failure for head and neck squamous cell carcinoma (HNSCC) patients. Identification of high risk subvolumes based on pretreatment imaging is key to biologically targeted radiation therapy. We investigated the extent to which a Convolutional neural network (CNN) is able to predict LRR volumes based on pre-treatment 18F-fluorodeoxyglucose positron emission tomography (FDG-PET)/computed tomography (CT) scans in HNSCC patients and thus the potential to identify biological high risk volumes using CNNs. For 37 patients who had undergone primary radiotherapy for oropharyngeal squamous cell carcinoma, five oncologists contoured the relapse volumes on recurrence CT scans. Datasets of pre-treatment FDG-PET/CT, gross tumour volume (GTV) and contoured relapse for each of the patients were randomly divided into training (n=23), validation (n=7) and test (n=7) datasets. We compared a CNN trained from scratch, a pre-trained CNN, a SUVmax threshold approach, and using the GTV directly. The SUVmax threshold method included 5 out of the 7 relapse origin points within a volume of median 4.6 cubic centimetres (cc). Both the GTV contour and best CNN segmentations included the relapse origin 6 out of 7 times with median volumes of 28 and 18 cc respectively. The CNN included the same or greater number of relapse volume POs, with significantly smaller relapse volumes. Our novel findings indicate that CNNs may predict LRR, yet further work on dataset development is required to attain clinically useful prediction accuracy.

Noor Fathima Ghouse, Jens Petersen, Auke Wiggers et al.

Diffusion probabilistic models have recently achieved remarkable success in generating high quality image and video data. In this work, we build on this class of generative models and introduce a method for lossy compression of high resolution images. The resulting codec, which we call DIffuson-based Residual Augmentation Codec (DIRAC), is the first neural codec to allow smooth traversal of the rate-distortion-perception tradeoff at test time, while obtaining competitive performance with GAN-based methods in perceptual quality. Furthermore, while sampling from diffusion probabilistic models is notoriously expensive, we show that in the compression setting the number of steps can be drastically reduced.

Shuhan Xiao, Lukas Klein, Jens Petersen et al.

Identifying predictive covariates, which forecast individual treatment effectiveness, is crucial for decision-making across different disciplines such as personalized medicine. These covariates, referred to as biomarkers, are extracted from pre-treatment data, often within randomized controlled trials, and should be distinguished from prognostic biomarkers, which are independent of treatment assignment. Our study focuses on discovering predictive imaging biomarkers, specific image features, by leveraging pre-treatment images to uncover new causal relationships. Unlike labor-intensive approaches relying on handcrafted features prone to bias, we present a novel task of directly learning predictive features from images. We propose an evaluation protocol to assess a model's ability to identify predictive imaging biomarkers and differentiate them from purely prognostic ones by employing statistical testing and a comprehensive analysis of image feature attribution. We explore the suitability of deep learning models originally developed for estimating the conditional average treatment effect (CATE) for this task, which have been assessed primarily for their precision of CATE estimation while overlooking the evaluation of imaging biomarker discovery. Our proof-of-concept analysis demonstrates the feasibility and potential of our approach in discovering and validating predictive imaging biomarkers from synthetic outcomes and real-world image datasets. Our code is available at \url{https://github.com/MIC-DKFZ/predictive_image_biomarker_analysis}.

Martina Finocchiaro, Ronja Stern, Abraham George Smith et al.

High-resolution colon segmentation is crucial for clinical and research applications, such as digital twins and personalized medicine. However, the leading open-source abdominal segmentation tool, TotalSegmentator, struggles with accuracy for the colon, which has a complex and variable shape, requiring time-intensive labeling. Here, we present the first fully automatic high-resolution colon segmentation method. To develop it, we first created a high resolution colon dataset using a pipeline that combines region growing with interactive machine learning to efficiently and accurately label the colon on CT colonography (CTC) images. Based on the generated dataset consisting of 435 labeled CTC images we trained an nnU-Net model for fully automatic colon segmentation. Our fully automatic model achieved an average symmetric surface distance of 0.2 mm (vs. 4.0 mm from TotalSegmentator) and a 95th percentile Hausdorff distance of 1.0 mm (vs. 18 mm from TotalSegmentator). Our segmentation accuracy substantially surpasses TotalSegmentator. We share our trained model and pipeline code, providing the first and only open-source tool for high-resolution colon segmentation. Additionally, we created a large-scale dataset of publicly available high-resolution colon labels.

Abraham George Smith, Jens Petersen, Cynthia Terrones-Campos et al.

Organ-at-risk contouring is still a bottleneck in radiotherapy, with many deep learning methods falling short of promised results when evaluated on clinical data. We investigate the accuracy and time-savings resulting from the use of an interactive-machine-learning method for an organ-at-risk contouring task. We compare the method to the Eclipse contouring software and find strong agreement with manual delineations, with a dice score of 0.95. The annotations created using corrective-annotation also take less time to create as more images are annotated, resulting in substantial time savings compared to manual methods, with hearts that take 2 minutes and 2 seconds to delineate on average, after 923 images have been delineated, compared to 7 minutes and 1 seconds when delineating manually. Our experiment demonstrates that interactive-machine-learning with corrective-annotation provides a fast and accessible way for non computer-scientists to train deep-learning models to segment their own structures of interest as part of routine clinical workflows. Source code is available at \href{https://github.com/Abe404/RootPainter3D}{this HTTPS URL}.

Fabian Isensee, Paul F. Jäger, Simon A. A. Kohl et al.

Biomedical imaging is a driver of scientific discovery and core component of medical care, currently stimulated by the field of deep learning. While semantic segmentation algorithms enable 3D image analysis and quantification in many applications, the design of respective specialised solutions is non-trivial and highly dependent on dataset properties and hardware conditions. We propose nnU-Net, a deep learning framework that condenses the current domain knowledge and autonomously takes the key decisions required to transfer a basic architecture to different datasets and segmentation tasks. Without manual tuning, nnU-Net surpasses most specialised deep learning pipelines in 19 public international competitions and sets a new state of the art in the majority of the 49 tasks. The results demonstrate a vast hidden potential in the systematic adaptation of deep learning methods to different datasets. We make nnU-Net publicly available as an open-source tool that can effectively be used out-of-the-box, rendering state of the art segmentation accessible to non-experts and catalyzing scientific progress as a framework for automated method design.

Amirhossein Habibian, Amir Ghodrati, Noor Fathima et al.

This work aims to improve the efficiency of text-to-image diffusion models. While diffusion models use computationally expensive UNet-based denoising operations in every generation step, we identify that not all operations are equally relevant for the final output quality. In particular, we observe that UNet layers operating on high-res feature maps are relatively sensitive to small perturbations. In contrast, low-res feature maps influence the semantic layout of the final image and can often be perturbed with no noticeable change in the output. Based on this observation, we propose Clockwork Diffusion, a method that periodically reuses computation from preceding denoising steps to approximate low-res feature maps at one or more subsequent steps. For multiple baselines, and for both text-to-image generation and image editing, we demonstrate that Clockwork leads to comparable or improved perceptual scores with drastically reduced computational complexity. As an example, for Stable Diffusion v1.5 with 8 DPM++ steps we save 32% of FLOPs with negligible FID and CLIP change.

Mohamed Omran, Dimitris Kalatzis, Jens Petersen et al.

Image editing approaches have become more powerful and flexible with the advent of powerful text-conditioned generative models. However, placing objects in an environment with a precise location and orientation still remains a challenge, as this typically requires carefully crafted inpainting masks or prompts. In this work, we show that a carefully designed visual map, combined with coarse object masks, is sufficient for high quality object placement. We design a conditioning signal that resolves ambiguities, while being flexible enough to allow for changing of shapes or object orientations. By building on an inpainting model, we leave the background intact by design, in contrast to methods that model objects and background jointly. We demonstrate the effectiveness of our method in the automotive setting, where we compare different conditioning signals in novel object placement tasks. These tasks are designed to measure edit quality not only in terms of appearance, but also in terms of pose and location accuracy, including cases that require non-trivial shape changes. Lastly, we show that fine location control can be combined with appearance control to place existing objects in precise locations in a scene.

Farhad G. Zanjani, Davide Abati, Auke Wiggers et al.

We investigate data augmentation for 3D object detection in autonomous driving. We utilize recent advancements in 3D reconstruction based on Gaussian Splatting for 3D object placement in driving scenes. Unlike existing diffusion-based methods that synthesize images conditioned on BEV layouts, our approach places 3D objects directly in the reconstructed 3D space with explicitly imposed geometric transformations. This ensures both the physical plausibility of object placement and highly accurate 3D pose and position annotations. Our experiments demonstrate that even by integrating a limited number of external 3D objects into real scenes, the augmented data significantly enhances 3D object detection performance and outperforms existing diffusion-based 3D augmentation for object detection. Extensive testing on the nuScenes dataset reveals that imposing high geometric diversity in object placement has a greater impact compared to the appearance diversity of objects. Additionally, we show that generating hard examples, either by maximizing detection loss or imposing high visual occlusion in camera images, does not lead to more efficient 3D data augmentation for camera-based 3D object detection in autonomous driving.

Mohamed Omran, Farhad Zanjani, Davide Abati et al.

This paper describes the Qualcomm AI Research solution to the RealADSim-NVS challenge, hosted at the RealADSim Workshop at ICCV 2025. The challenge concerns novel view synthesis in street scenes, and participants are required to generate, starting from car-centric frames captured during some training traversals, renders of the same urban environment as viewed from a different traversal (e.g. different street lane or car direction). Our solution is inspired by hybrid methods in scene generation and generative simulators merging gaussian splatting and diffusion models, and it is composed of two stages: First, we fit a 3D reconstruction of the scene and render novel views as seen from the target cameras. Then, we enhance the resulting frames with a dedicated single-step diffusion model. We discuss specific choices made in the initialization of gaussian primitives as well as the finetuning of the enhancer model and its training data curation. We report the performance of our model design and we ablate its components in terms of novel view quality as measured by PSNR, SSIM and LPIPS. On the public leaderboard reporting test results, our proposal reaches an aggregated score of 0.432, achieving the second place overall.

Emily Honey, Anders Helbo, Jens Petersen

Computed tomography (CT) is essential for treatment and diagnostics; In case CT are missing or otherwise difficult to obtain, methods for generating synthetic CT (sCT) images from magnetic resonance imaging (MRI) images are sought after. Therefore, it is valuable to establish a reference for what strategies are most effective for MRI-to-CT translation. In this paper, we compare the performance of two frequently used architectures for MRI-to-CT translation: a conditional generative adversarial network (cGAN) and a conditional denoising diffusion probabilistic model (cDDPM). We chose well-established implementations to represent each architecture: Pix2Pix for cGAN, and Palette for cDDPM. We separate the classical 3D translation problem into a sequence of 2D translations on the transverse plane, to investigate the viability of a strategy that reduces the computational cost. We also investigate the impact of conditioning the generative process on a single MRI image/slice and on multiple MRI slices. The performance is assessed using a thorough evaluation protocol, including a novel slice-wise metric Similarity Of Slices (SIMOS), which measures the continuity between transverse slices when compiling the sCTs into 3D format. Our comparative analysis revealed that MRI-to-CT generative models benefit from multi-channel conditional input and using cDDPM as an architecture.

Silas Nyboe Ørting, Kristina Miger, Anne Sophie Overgaard Olesen et al.

Introduction: Chest CT scans are increasingly used in dyspneic patients where acute heart failure (AHF) is a key differential diagnosis. Interpretation remains challenging and radiology reports are frequently delayed due to a radiologist shortage, although flagging such information for emergency physicians would have therapeutic implication. Artificial intelligence (AI) can be a complementary tool to enhance the diagnostic precision. We aim to develop an explainable AI model to detect radiological signs of AHF in chest CT with an accuracy comparable to thoracic radiologists. Methods: A single-center, retrospective study during 2016-2021 at Copenhagen University Hospital - Bispebjerg and Frederiksberg, Denmark. A Boosted Trees model was trained to predict AHF based on measurements of segmented cardiac and pulmonary structures from acute thoracic CT scans. Diagnostic labels for training and testing were extracted from radiology reports. Structures were segmented with TotalSegmentator. Shapley Additive explanations values were used to explain the impact of each measurement on the final prediction. Results: Of the 4,672 subjects, 49% were female. The final model incorporated twelve key features of AHF and achieved an area under the ROC of 0.87 on the independent test set. Expert radiologist review of model misclassifications found that 24 out of 64 (38%) false positives and 24 out of 61 (39%) false negatives were actually correct model predictions, with the errors originating from inaccuracies in the initial radiology reports. Conclusion: We developed an explainable AI model with strong discriminatory performance, comparable to thoracic radiologists. The AI model's stepwise, transparent predictions may support decision-making.

Jens Petersen, Davide Abati, Amirhossein Habibian et al.

Generative image models are increasingly being used for training data augmentation in vision tasks. In the context of automotive object detection, methods usually focus on producing augmented frames that look as realistic as possible, for example by replacing real objects with generated ones. Others try to maximize the diversity of augmented frames, for example by pasting lots of generated objects onto existing backgrounds. Both perspectives pay little attention to the locations of objects in the scene. Frame layouts are either reused with little or no modification, or they are random and disregard realism entirely. In this work, we argue that optimal data augmentation should also include realistic augmentation of layouts. We introduce a scene-aware probabilistic location model that predicts where new objects can realistically be placed in an existing scene. By then inpainting objects in these locations with a generative model, we obtain much stronger augmentation performance than existing approaches. We set a new state of the art for generative data augmentation on two automotive object detection tasks, achieving up to $2.8\times$ higher gains than the best competing approach ($+1.4$ vs. $+0.5$ mAP boost). We also demonstrate significant improvements for instance segmentation.

Raghavendra Selvan, Erik B Dam, Søren Alexander Flensborg et al.

Tensor networks are efficient factorisations of high-dimensional tensors into a network of lower-order tensors. They have been most commonly used to model entanglement in quantum many-body systems and more recently are witnessing increased applications in supervised machine learning. In this work, we formulate image segmentation in a supervised setting with tensor networks. The key idea is to first lift the pixels in image patches to exponentially high-dimensional feature spaces and using a linear decision hyper-plane to classify the input pixels into foreground and background classes. The high-dimensional linear model itself is approximated using the matrix product state (MPS) tensor network. The MPS is weight-shared between the non-overlapping image patches resulting in our strided tensor network model. The performance of the proposed model is evaluated on three 2D- and one 3D- biomedical imaging datasets. The performance of the proposed tensor network segmentation model is compared with relevant baseline methods. In the 2D experiments, the tensor network model yields competitive performance compared to the baseline methods while being more resource efficient.

Jens Petersen, Fabian Isensee, Gregor Köhler et al.

The ability to estimate how a tumor might evolve in the future could have tremendous clinical benefits, from improved treatment decisions to better dose distribution in radiation therapy. Recent work has approached the glioma growth modeling problem via deep learning and variational inference, thus learning growth dynamics entirely from a real patient data distribution. So far, this approach was constrained to predefined image acquisition intervals and sequences of fixed length, which limits its applicability in more realistic scenarios. We overcome these limitations by extending Neural Processes, a class of conditional generative models for stochastic time series, with a hierarchical multi-scale representation encoding including a spatio-temporal attention mechanism. The result is a learned growth model that can be conditioned on an arbitrary number of observations, and that can produce a distribution of temporally consistent growth trajectories on a continuous time axis. On a dataset of 379 patients, the approach successfully captures both global and finer-grained variations in the images, exhibiting superior performance compared to other learned growth models.

Jens Petersen, Gregor Köhler, David Zimmerer et al.

Neural Processes (NPs) are a family of conditional generative models that are able to model a distribution over functions, in a way that allows them to perform predictions at test time conditioned on a number of context points. A recent addition to this family, Convolutional Conditional Neural Processes (ConvCNP), have shown remarkable improvement in performance over prior art, but we find that they sometimes struggle to generalize when applied to time series data. In particular, they are not robust to distribution shifts and fail to extrapolate observed patterns into the future. By incorporating a Gaussian Process into the model, we are able to remedy this and at the same time improve performance within distribution. As an added benefit, the Gaussian Process reintroduces the possibility to sample from the model, a key feature of other members in the NP family.

Annika Reinke, Minu D. Tizabi, Carole H. Sudre et al.

While the importance of automatic image analysis is continuously increasing, recent meta-research revealed major flaws with respect to algorithm validation. Performance metrics are particularly key for meaningful, objective, and transparent performance assessment and validation of the used automatic algorithms, but relatively little attention has been given to the practical pitfalls when using specific metrics for a given image analysis task. These are typically related to (1) the disregard of inherent metric properties, such as the behaviour in the presence of class imbalance or small target structures, (2) the disregard of inherent data set properties, such as the non-independence of the test cases, and (3) the disregard of the actual biomedical domain interest that the metrics should reflect. This living dynamically document has the purpose to illustrate important limitations of performance metrics commonly applied in the field of image analysis. In this context, it focuses on biomedical image analysis problems that can be phrased as image-level classification, semantic segmentation, instance segmentation, or object detection task. The current version is based on a Delphi process on metrics conducted by an international consortium of image analysis experts from more than 60 institutions worldwide.

Raghavendra Selvan, Erik B Dam, Jens Petersen

Tensor networks provide an efficient approximation of operations involving high dimensional tensors and have been extensively used in modelling quantum many-body systems. More recently, supervised learning has been attempted with tensor networks, primarily focused on tasks such as image classification. In this work, we propose a novel formulation of tensor networks for supervised image segmentation which allows them to operate on high resolution medical images. We use the matrix product state (MPS) tensor network on non-overlapping patches of a given input image to predict the segmentation mask by learning a pixel-wise linear classification rule in a high dimensional space. The proposed model is end-to-end trainable using backpropagation. It is implemented as a Strided Tensor Network to reduce the parameter complexity. The performance of the proposed method is evaluated on two public medical imaging datasets and compared to relevant baselines. The evaluation shows that the strided tensor network yields competitive performance compared to CNN-based models while using fewer resources. Additionally, based on the experiments we discuss the feasibility of using fully linear models for segmentation tasks.

David Zimmerer, Jens Petersen, Simon A. A. Kohl et al.

Through training on unlabeled data, anomaly detection has the potential to impact computer-aided diagnosis by outlining suspicious regions. Previous work on deep-learning-based anomaly detection has primarily focused on the reconstruction error. We argue instead, that pixel-wise anomaly ratings derived from a Variational Autoencoder based score approximation yield a theoretically better grounded and more faithful estimate. In our experiments, Variational Autoencoder gradient-based rating outperforms other approaches on unsupervised pixel-wise tumor detection on the BraTS-2017 dataset with a ROC-AUC of 0.94.

David Zimmerer, Jens Petersen, Klaus Maier-Hein

Variational Auto-Encoders have often been used for unsupervised pretraining, feature extraction and out-of-distribution and anomaly detection in the medical field. However, VAEs often lack the ability to produce sharp images and learn high-level features. We propose to alleviate these issues by adding a new branch to conditional hierarchical VAEs. This enforces a division between higher-level and lower-level features. Despite the additional computational overhead compared to a normal VAE it results in sharper and better reconstructions and can capture the data distribution similarly well (indicated by a similar or slightly better OoD detection performance).

Jens Petersen, Paul F. Jäger, Fabian Isensee et al.

Existing approaches to modeling the dynamics of brain tumor growth, specifically glioma, employ biologically inspired models of cell diffusion, using image data to estimate the associated parameters. In this work, we propose an alternative approach based on recent advances in probabilistic segmentation and representation learning that implicitly learns growth dynamics directly from data without an underlying explicit model. We present evidence that our approach is able to learn a distribution of plausible future tumor appearances conditioned on past observations of the same tumor.

David Zimmerer, Fabian Isensee, Jens Petersen et al.

An assumption-free automatic check of medical images for potentially overseen anomalies would be a valuable assistance for a radiologist. Deep learning and especially Variational Auto-Encoders (VAEs) have shown great potential in the unsupervised learning of data distributions. In principle, this allows for such a check and even the localization of parts in the image that are most suspicious. Currently, however, the reconstruction-based localization by design requires adjusting the model architecture to the specific problem looked at during evaluation. This contradicts the principle of building assumption-free models. We propose complementing the localization part with a term derived from the Kullback-Leibler (KL)-divergence. For validation, we perform a series of experiments on FashionMNIST as well as on a medical task including >1000 healthy and >250 brain tumor patients. Results show that the proposed formalism outperforms the state of the art VAE-based localization of anomalies across many hyperparameter settings and also shows a competitive max performance.

Wouter M. Kouw, Silas N. Ørting, Jens Petersen et al.

Suppose one is faced with the challenge of tissue segmentation in MR images, without annotators at their center to provide labeled training data. One option is to go to another medical center for a trained classifier. Sadly, tissue classifiers do not generalize well across centers due to voxel intensity shifts caused by center-specific acquisition protocols. However, certain aspects of segmentations, such as spatial smoothness, remain relatively consistent and can be learned separately. Here we present a smoothness prior that is fit to segmentations produced at another medical center. This informative prior is presented to an unsupervised Bayesian model. The model clusters the voxel intensities, such that it produces segmentations that are similarly smooth to those of the other medical center. In addition, the unsupervised Bayesian model is extended to a semi-supervised variant, which needs no visual interpretation of clusters into tissues.

Abraham George Smith, Jens Petersen, Raghavendra Selvan et al.

Plant root research can provide a way to attain stress-tolerant crops that produce greater yield in a diverse array of conditions. Phenotyping roots in soil is often challenging due to the roots being difficult to access and the use of time consuming manual methods. Rhizotrons allow visual inspection of root growth through transparent surfaces. Agronomists currently manually label photographs of roots obtained from rhizotrons using a line-intersect method to obtain root length density and rooting depth measurements which are essential for their experiments. We investigate the effectiveness of an automated image segmentation method based on the U-Net Convolutional Neural Network (CNN) architecture to enable such measurements. We design a data-set of 50 annotated Chicory (Cichorium intybus L.) root images which we use to train, validate and test the system and compare against a baseline built using the Frangi vesselness filter. We obtain metrics using manual annotations and line-intersect counts. Our results on the held out data show our proposed automated segmentation system to be a viable solution for detecting and quantifying roots. We evaluate our system using 867 images for which we have obtained line-intersect counts, attaining a Spearman rank correlation of 0.9748 and an $r^2$ of 0.9217. We also achieve an $F_1$ of 0.7 when comparing the automated segmentation to the manual annotations, with our automated segmentation system producing segmentations with higher quality than the manual annotations for large portions of the image.

Patrick Bilic, Patrick Christ, Hongwei Bran Li et al.

In this work, we report the set-up and results of the Liver Tumor Segmentation Benchmark (LiTS), which was organized in conjunction with the IEEE International Symposium on Biomedical Imaging (ISBI) 2017 and the International Conferences on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2017 and 2018. The image dataset is diverse and contains primary and secondary tumors with varied sizes and appearances with various lesion-to-background levels (hyper-/hypo-dense), created in collaboration with seven hospitals and research institutions. Seventy-five submitted liver and liver tumor segmentation algorithms were trained on a set of 131 computed tomography (CT) volumes and were tested on 70 unseen test images acquired from different patients. We found that not a single algorithm performed best for both liver and liver tumors in the three events. The best liver segmentation algorithm achieved a Dice score of 0.963, whereas, for tumor segmentation, the best algorithms achieved Dices scores of 0.674 (ISBI 2017), 0.702 (MICCAI 2017), and 0.739 (MICCAI 2018). Retrospectively, we performed additional analysis on liver tumor detection and revealed that not all top-performing segmentation algorithms worked well for tumor detection. The best liver tumor detection method achieved a lesion-wise recall of 0.458 (ISBI 2017), 0.515 (MICCAI 2017), and 0.554 (MICCAI 2018), indicating the need for further research. LiTS remains an active benchmark and resource for research, e.g., contributing the liver-related segmentation tasks in \url{http://medicaldecathlon.com/}. In addition, both data and online evaluation are accessible via \url{www.lits-challenge.com}.

David Zimmerer, Simon A. A. Kohl, Jens Petersen et al.

Unsupervised learning can leverage large-scale data sources without the need for annotations. In this context, deep learning-based auto encoders have shown great potential in detecting anomalies in medical images. However, state-of-the-art anomaly scores are still based on the reconstruction error, which lacks in two essential parts: it ignores the model-internal representation employed for reconstruction, and it lacks formal assertions and comparability between samples. We address these shortcomings by proposing the Context-encoding Variational Autoencoder (ceVAE) which combines reconstruction- with density-based anomaly scoring. This improves the sample- as well as pixel-wise results. In our experiments on the BraTS-2017 and ISLES-2015 segmentation benchmarks, the ceVAE achieves unsupervised ROC-AUCs of 0.95 and 0.89, respectively, thus outperforming state-of-the-art methods by a considerable margin.

Raghavendra Selvan, Thomas Kipf, Max Welling et al.

Graph refinement, or the task of obtaining subgraphs of interest from over-complete graphs, can have many varied applications. In this work, we extract trees or collection of sub-trees from image data by, first deriving a graph-based representation of the volumetric data and then, posing the tree extraction as a graph refinement task. We present two methods to perform graph refinement. First, we use mean-field approximation (MFA) to approximate the posterior density over the subgraphs from which the optimal subgraph of interest can be estimated. Mean field networks (MFNs) are used for inference based on the interpretation that iterations of MFA can be seen as feed-forward operations in a neural network. This allows us to learn the model parameters using gradient descent. Second, we present a supervised learning approach using graph neural networks (GNNs) which can be seen as generalisations of MFNs. Subgraphs are obtained by training a GNN-based graph refinement model to directly predict edge probabilities. We discuss connections between the two classes of methods and compare them for the task of extracting airways from 3D, low-dose, chest CT data. We show that both the MFN and GNN models show significant improvement when compared to one baseline method, that is similar to a top performing method in the EXACT'09 Challenge, and a 3D U-Net based airway segmentation model, in detecting more branches with fewer false positives.

Silas Nyboe Ørting, Jens Petersen, Laura H. Thomsen et al.

Accurate assessment of pulmonary emphysema is crucial to assess disease severity and subtype, to monitor disease progression and to predict lung cancer risk. However, visual assessment is time-consuming and subject to substantial inter-rater variability and standard densitometry approaches to quantify emphysema remain inferior to visual scoring. We explore if machine learning methods that learn from a large dataset of visually assessed CT scans can provide accurate estimates of emphysema extent. We further investigate if machine learning algorithms that learn from a scoring of emphysema extent can outperform algorithms that learn only from a scoring of emphysema presence. We compare four Multiple Instance Learning classifiers that are trained on emphysema presence labels, and five Learning with Label Proportions classifiers that are trained on emphysema extent labels. We evaluate performance on 600 low-dose CT scans from the Danish Lung Cancer Screening Trial and find that learning from emphysema presence labels, which are much easier to obtain, gives equally good performance to learning from emphysema extent labels. The best classifiers achieve intra-class correlation coefficients around 0.90 and average overall agreement with raters of 78% and 79% on six emphysema extent classes versus inter-rater agreement of 83%.

Fabian Isensee, Jens Petersen, Andre Klein et al.

The U-Net was presented in 2015. With its straight-forward and successful architecture it quickly evolved to a commonly used benchmark in medical image segmentation. The adaptation of the U-Net to novel problems, however, comprises several degrees of freedom regarding the exact architecture, preprocessing, training and inference. These choices are not independent of each other and substantially impact the overall performance. The present paper introduces the nnU-Net ('no-new-Net'), which refers to a robust and self-adapting framework on the basis of 2D and 3D vanilla U-Nets. We argue the strong case for taking away superfluous bells and whistles of many proposed network designs and instead focus on the remaining aspects that make out the performance and generalizability of a method. We evaluate the nnU-Net in the context of the Medical Segmentation Decathlon challenge, which measures segmentation performance in ten disciplines comprising distinct entities, image modalities, image geometries and dataset sizes, with no manual adjustments between datasets allowed. At the time of manuscript submission, nnU-Net achieves the highest mean dice scores across all classes and seven phase 1 tasks (except class 1 in BrainTumour) in the online leaderboard of the challenge.

Raghavendra Selvan, Jens Petersen, Jesper H Pedersen et al.

In this work, we adapt a method based on multiple hypothesis tracking (MHT) that has been shown to give state-of-the-art vessel segmentation results in interactive settings, for the purpose of extracting trees. Regularly spaced tubular templates are fit to image data forming local hypotheses. These local hypotheses are used to construct the MHT tree, which is then traversed to make segmentation decisions. However, some critical parameters in this method are scale-dependent and have an adverse effect when tracking structures of varying dimensions. We propose to use statistical ranking of local hypotheses in constructing the MHT tree, which yields a probabilistic interpretation of scores across scales and helps alleviate the scale-dependence of MHT parameters. This enables our method to track trees starting from a single seed point. Our method is evaluated on chest CT data to extract airway trees and coronary arteries. In both cases, we show that our method performs significantly better than the original MHT method.

Silas Nyboe Ørting, Jens Petersen, Veronika Cheplygina et al.

Supervised feature learning using convolutional neural networks (CNNs) can provide concise and disease relevant representations of medical images. However, training CNNs requires annotated image data. Annotating medical images can be a time-consuming task and even expert annotations are subject to substantial inter- and intra-rater variability. Assessing visual similarity of images instead of indicating specific pathologies or estimating disease severity could allow non-experts to participate, help uncover new patterns, and possibly reduce rater variability. We consider the task of assessing emphysema extent in chest CT scans. We derive visual similarity triplets from visually assessed emphysema extent and learn a low dimensional embedding using CNNs. We evaluate the networks on 973 images, and show that the CNNs can learn disease relevant feature representations from derived similarity triplets. To our knowledge this is the first medical image application where similarity triplets has been used to learn a feature representation that can be used for embedding unseen test images

Raghavendra Selvan, Thomas Kipf, Max Welling et al.

We present extraction of tree structures, such as airways, from image data as a graph refinement task. To this end, we propose a graph auto-encoder model that uses an encoder based on graph neural networks (GNNs) to learn embeddings from input node features and a decoder to predict connections between nodes. Performance of the GNN model is compared with mean-field networks in their ability to extract airways from 3D chest CT scans.

Raghavendra Selvan, Max Welling, Jesper H. Pedersen et al.

We present tree extraction in 3D images as a graph refinement task, of obtaining a subgraph from an over-complete input graph. To this end, we formulate an approximate Bayesian inference framework on undirected graphs using mean field approximation (MFA). Mean field networks are used for inference based on the interpretation that iterations of MFA can be seen as feed-forward operations in a neural network. This allows us to learn the model parameters from training data using back-propagation algorithm. We demonstrate usefulness of the model to extract airway trees from 3D chest CT data. We first obtain probability images using a voxel classifier that distinguishes airways from background and use Bayesian smoothing to model individual airway branches. This yields us joint Gaussian density estimates of position, orientation and scale as node features of the input graph. Performance of the method is compared with two methods: the first uses probability images from a trained voxel classifier with region growing, which is similar to one of the best performing methods at EXACT'09 airway challenge, and the second method is based on Bayesian smoothing on these probability images. Using centerline distance as error measure the presented method shows significant improvement compared to these two methods.

Raghavendra Selvan, Jens Petersen, Jesper H. Pedersen et al.

Segmenting tree structures is common in several image processing applications. In medical image analysis, reliable segmentations of airways, vessels, neurons and other tree structures can enable important clinical applications. We present a framework for tracking tree structures comprising of elongated branches using probabilistic state-space models and Bayesian smoothing. Unlike most existing methods that proceed with sequential tracking of branches, we present an exploratory method, that is less sensitive to local anomalies in the data due to acquisition noise and/or interfering structures. The evolution of individual branches is modelled using a process model and the observed data is incorporated into the update step of the Bayesian smoother using a measurement model that is based on a multi-scale blob detector. Bayesian smoothing is performed using the RTS (Rauch-Tung-Striebel) smoother, which provides Gaussian density estimates of branch states at each tracking step. We select likely branch seed points automatically based on the response of the blob detection and track from all such seed points using the RTS smoother. We use covariance of the marginal posterior density estimated for each branch to discriminate false positive and true positive branches. The method is evaluated on 3D chest CT scans to track airways. We show that the presented method results in additional branches compared to a baseline method based on region growing on probability images.

Raghavendra Selvan, Jens Petersen, Jesper H. Pedersen et al.

Knowledge of airway tree morphology has important clinical applications in diagnosis of chronic obstructive pulmonary disease. We present an automatic tree extraction method based on multiple hypothesis tracking and template matching for this purpose and evaluate its performance on chest CT images. The method is adapted from a semi-automatic method devised for vessel segmentation. Idealized tubular templates are constructed that match airway probability obtained from a trained classifier and ranked based on their relative significance. Several such regularly spaced templates form the local hypotheses used in constructing a multiple hypothesis tree, which is then traversed to reach decisions. The proposed modifications remove the need for local thresholding of hypotheses as decisions are made entirely based on statistical comparisons involving the hypothesis tree. The results show improvements in performance when compared to the original method and region growing on intensity images. We also compare the method with region growing on the probability images, where the presented method does not show substantial improvement, but we expect it to be less sensitive to local anomalies in the data.

Aasa Feragen, Jens Petersen, Dominik Grimm et al.

Methodological contributions: This paper introduces a family of kernels for analyzing (anatomical) trees endowed with vector valued measurements made along the tree. While state-of-the-art graph and tree kernels use combinatorial tree/graph structure with discrete node and edge labels, the kernels presented in this paper can include geometric information such as branch shape, branch radius or other vector valued properties. In addition to being flexible in their ability to model different types of attributes, the presented kernels are computationally efficient and some of them can easily be computed for large datasets (N of the order 10.000) of trees with 30-600 branches. Combining the kernels with standard machine learning tools enables us to analyze the relation between disease and anatomical tree structure and geometry. Experimental results: The kernels are used to compare airway trees segmented from low-dose CT, endowed with branch shape descriptors and airway wall area percentage measurements made along the tree. Using kernelized hypothesis testing we show that the geometric airway trees are significantly differently distributed in patients with Chronic Obstructive Pulmonary Disease (COPD) than in healthy individuals. The geometric tree kernels also give a significant increase in the classification accuracy of COPD from geometric tree structure endowed with airway wall thickness measurements in comparison with state-of-the-art methods, giving further insight into the relationship between airway wall thickness and COPD. Software: Software for computing kernels and statistical tests is available at http://image.diku.dk/aasa/software.php.