Cone-beam computed tomography (CBCT) is widely used for dento-maxillofacial diagnostics and treatment planning, and comprehensive multi-structure segmentation remains time-consuming, limiting large-scale, reproducible research. In this article, we present ToothFairy2, a MICCAI 2024 challenge on multi-structure segmentation in maxillofacial CBCT. The accompanying dataset comprises 530 CBCT volumes (480 public training, 50 hidden test) with expert 3D annotations of 42 classes, including maxilla, mandible, crowns, bridges, implants, inferior alveolar canals, maxillary sinuses, pharynx, and teeth labeled according to the International Tooth Numbering System (FDI). 26 international teams participated in ToothFairy2, and their methods were run and evaluated for voxel-wise multi-class segmentation using a standardized protocol. This report extends the evaluation of teeth to also investigate the current capabilities of tooth detection and FDI numbering. Furthermore, ranking stability was analyzed to assess the robustness of the final challenge outcome. Overall, challenge participants achieved consistently high performance for large, high-contrast structures such as jawbones, pharynx, and most teeth, while maxillary sinuses, dental restorations, and fine structures remain challenging due to class imbalance and metal artifacts. Analysis of tooth-related metrics further revealed that assigning correct FDI numbers was more challenging than delineating individual teeth. By releasing CBCT data, 3D annotations, baseline models, and evaluation code, ToothFairy2 establishes a long-term benchmark to drive the development of automated methods for robust, clinically meaningful multi-structure segmentation in maxillofacial CBCT.

Atrial fibrillation (AF), the most common cardiac arrhythmia, affects one in three adults over 45 years of age. Improving its treatment requires a better understanding of bi-atrial anatomy. Existing benchmarks have focused on the left atrial (LA) cavity, overlooking the fundamental challenges posed by bi-atrial anatomy, most notably the thin atrial walls, which are critical for substrate-guided ablation planning in patients with atrial fibrillation. To address these limitations, the Multi-class Bi-Atrial Segmentation 2024 Challenge (MBAS2024) introduced the first large-scale, multi-class benchmark for simultaneous segmentation of the LA cavity, right atrial (RA) cavity, and bi-atrial walls from late gadolinium-enhanced (LGE) MRI. We systematically evaluated 13 state-of-the-art methods on the world's largest curated bi-atrial dataset, comprising 175 3D multi-center scans with expert-validated annotations, providing a comprehensive assessment of current methodological capabilities and limitations. Key findings include: segmentation of the LA and RA cavities is generally robust to image quality, whereas atrial wall delineation is highly sensitive to image degradation. Performance varies across centers, indicating limited generalization of atrial wall segmentation across different acquisition protocols. Model architecture, rather than hyperparameter tuning, is the primary driver of performance, with U-Net-based models and emerging state-space models (e.g., UMambaBot) achieving higher accuracy at modest computational cost. Segmentation accuracy also varies along the slice dimension, with central slices segmented more reliably. Finally, hybrid labeling strategies-separating LA and RA cavities while merging bi-atrial walls into a single class-consistently improve performance. The MBAS2024 challenge establishes a foundational benchmark for bi-atrial segmentation, providing validated baselines and actionable insights to guide the development of clinically relevant, efficient, and anatomically aware segmentation algorithms to improve targeted ablation in patients with AF.

In this work, we present LesionLocator, a framework for zero-shot longitudinal lesion tracking and segmentation in 3D medical imaging, establishing the first end-to-end model capable of 4D tracking with dense spatial prompts. Our model leverages an extensive dataset of 23,262 annotated medical scans, as well as synthesized longitudinal data across diverse lesion types. The diversity and scale of our dataset significantly enhances model generalizability to real-world medical imaging challenges and addresses key limitations in longitudinal data availability. LesionLocator outperforms all existing promptable models in lesion segmentation by nearly 10 dice points, reaching human-level performance, and achieves state-of-the-art results in lesion tracking, with superior lesion retrieval and segmentation accuracy. LesionLocator not only sets a new benchmark in universal promptable lesion segmentation and automated longitudinal lesion tracking but also provides the first open-access solution of its kind, releasing our synthetic 4D dataset and model to the community, empowering future advancements in medical imaging. Code is available at: www.github.com/MIC-DKFZ/LesionLocator

Self-Supervised Learning (SSL) presents an exciting opportunity to unlock the potential of vast, untapped clinical datasets, for various downstream applications that suffer from the scarcity of labeled data. While SSL has revolutionized fields like natural language processing and computer vision, its adoption in 3D medical image computing has been limited by three key pitfalls: Small pre-training dataset sizes, architectures inadequate for 3D medical image analysis, and insufficient evaluation practices. In this paper, we address these issues by i) leveraging a large-scale dataset of 39k 3D brain MRI volumes and ii) using a Residual Encoder U-Net architecture within the state-of-the-art nnU-Net framework. iii) A robust development framework, incorporating 5 development and 8 testing brain MRI segmentation datasets, allowed performance-driven design decisions to optimize the simple concept of Masked Auto Encoders (MAEs) for 3D CNNs. The resulting model not only surpasses previous SSL methods but also outperforms the strong nnU-Net baseline by an average of approximately 3 Dice points setting a new state-of-the-art. Our code and models are made available here.

The field of self-supervised learning (SSL) for 3D medical images lacks consistency and standardization. While many methods have been developed, it is impossible to identify the current state-of-the-art, due to i) varying and small pretraining datasets, ii) varying architectures, and iii) being evaluated on differing downstream datasets. In this paper, we bring clarity to this field and lay the foundation for further method advancements through three key contributions: We a) publish the largest publicly available pre-training dataset comprising 114k 3D brain MRI volumes, enabling all practitioners to pre-train on a large-scale dataset. We b) benchmark existing 3D self-supervised learning methods on this dataset for a state-of-the-art CNN and Transformer architecture, clarifying the state of 3D SSL pre-training. Among many findings, we show that pre-trained methods can exceed a strong from-scratch nnU-Net ResEnc-L baseline. Lastly, we c) publish the code of our pre-training and fine-tuning frameworks and provide the pre-trained models created during the benchmarking process to facilitate rapid adoption and reproduction.

How can we test AI performance? This question seems trivial, but it isn't. Standard benchmarks often have problems such as in-distribution and small-size test sets, oversimplified metrics, unfair comparisons, and short-term outcome pressure. As a consequence, good performance on standard benchmarks does not guarantee success in real-world scenarios. To address these problems, we present Touchstone, a large-scale collaborative segmentation benchmark of 9 types of abdominal organs. This benchmark is based on 5,195 training CT scans from 76 hospitals around the world and 5,903 testing CT scans from 11 additional hospitals. This diverse test set enhances the statistical significance of benchmark results and rigorously evaluates AI algorithms across various out-of-distribution scenarios. We invited 14 inventors of 19 AI algorithms to train their algorithms, while our team, as a third party, independently evaluated these algorithms on three test sets. In addition, we also evaluated pre-existing AI frameworks---which, differing from algorithms, are more flexible and can support different algorithms—including MONAI from NVIDIA, nnU-Net from DKFZ, and numerous other open-source frameworks. We are committed to expanding this benchmark to encourage more innovation of AI algorithms for the medical domain.

What representation do deep neural networks learn? How similar are images to each other for neural networks? Despite the overwhelming success of deep learning methods key questions about their internal workings still remain largely unanswered, due to their internal high dimensionality and complexity. To address this, one approach is to measure the similarity of activation responses to various inputs. Representational Similarity Matrices (RSMs) distill this similarity into scalar values for each input pair. These matrices encapsulate the entire similarity structure of a system, indicating which input lead to similar responses. While the similarity between images is ambiguous, we argue that the spatial location of semantic objects does neither influence human perception nor deep learning classifiers. Thus this should be reflected in the definition of similarity between image responses for computer vision systems. Revisiting the established similarity calculations for RSMs we expose their sensitivity to spatial alignment. In this paper we propose to solve this through _semantic RSMs_, which are invariant to spatial permutation. We measure semantic similarity between input responses by formulating it as a set-matching problem. Further, we quantify the superiority of _semantic_ RSMs over _spatio-semantic_ RSMs through image retrieval and by comparing the similarity between representations to the similarity between predicted class probabilities.

Despite the remarkable success of deep learning systems over the last decade, a key difference still remains between neural network and human decision-making: As humans, we can not only form a decision on the spot, but also ponder, revisiting an initial guess from different angles, distilling relevant information, arriving at a better decision. Here, we propose RecycleNet, a latent feature recycling method, instilling the pondering capability for neural networks to refine initial decisions over a number of recycling steps, where outputs are fed back into earlier network layers in an iterative fashion. This approach makes minimal assumptions about the neural network architecture and thus can be implemented in a wide variety of contexts. Using medical image segmentation as the evaluation environment, we show that latent feature recycling enables the network to iteratively refine initial predictions even beyond the iterations seen during training, converging towards an improved decision. We evaluate this across a variety of segmentation benchmarks and show consistent improvements even compared with top-performing segmentation methods. This allows trading increased computation time for improved performance, which can be beneficial, especially for safety-critical applications.