Calibration and Uncertainty for multiRater Volume Assessment in multiorgan Segmentation (CURVAS) challenge results

Deep learning (DL) has become the dominant approach for medical image segmentation, yet ensuring the reliability and clinical applicability of these models requires addressing key challenges such as annotation variability, calibration, and uncertainty estimation. This is why we created the Calibration and Uncertainty for multiRater Volume Assessment in multiorgan Segmentation (CURVAS), which highlights the critical role of multiple annotators in establishing a more comprehensive ground truth, emphasizing that segmentation is inherently subjective and that leveraging inter-annotator variability is essential for robust model evaluation. Seven teams participated in the challenge, submitting a variety of DL models evaluated using metrics such as Dice Similarity Coefficient (DSC), Expected Calibration Error (ECE), and Continuous Ranked Probability Score (CRPS). By incorporating consensus and dissensus ground truth, we assess how DL models handle uncertainty and whether their confidence estimates align with true segmentation performance. Our findings reinforce the importance of well-calibrated models, as better calibration is strongly correlated with the quality of the results. Furthermore, we demonstrate that segmentation models trained on diverse datasets and enriched with pre-trained knowledge exhibit greater robustness, particularly in cases deviating from standard anatomical structures. Notably, the best-performing models achieved high DSC and well-calibrated uncertainty estimates. This work underscores the need for multi-annotator ground truth, thorough calibration assessments, and uncertainty-aware evaluations to develop trustworthy and clinically reliable DL-based medical image segmentation models.

Hydra: Marker-Free RGB-D Hand-Eye Calibration

This work presents an RGB-D imaging-based approach to marker-free hand-eye calibration using a novel implementation of the iterative closest point (ICP) algorithm with a robust point-to-plane (PTP) objective formulated on a Lie algebra. Its applicability is demonstrated through comprehensive experiments using three well known serial manipulators and two RGB-D cameras. With only three randomly chosen robot configurations, our approach achieves approximately 90% successful calibrations, demonstrating 2-3x higher convergence rates to the global optimum compared to both marker-based and marker-free baselines. We also report 2 orders of magnitude faster convergence time (0.8 +/- 0.4 s) for 9 robot configurations over other marker-free methods. Our method exhibits significantly improved accuracy (5 mm in task space) over classical approaches (7 mm in task space) whilst being marker-free. The benchmarking dataset and code are open sourced under Apache 2.0 License, and a ROS 2 integration with robot abstraction is provided to facilitate deployment.

Surgeons vs. Computer Vision: A comparative analysis on surgical phase recognition capabilities

Purpose: Automated Surgical Phase Recognition (SPR) uses Artificial Intelligence (AI) to segment the surgical workflow into its key events, functioning as a building block for efficient video review, surgical education as well as skill assessment. Previous research has focused on short and linear surgical procedures and has not explored if temporal context influences experts' ability to better classify surgical phases. This research addresses these gaps, focusing on Robot-Assisted Partial Nephrectomy (RAPN) as a highly non-linear procedure. Methods: Urologists of varying expertise were grouped and tasked to indicate the surgical phase for RAPN on both single frames and video snippets using a custom-made web platform. Participants reported their confidence levels and the visual landmarks used in their decision-making. AI architectures without and with temporal context as trained and benchmarked on the Cholec80 dataset were subsequently trained on this RAPN dataset. Results: Video snippets and presence of specific visual landmarks improved phase classification accuracy across all groups. Surgeons displayed high confidence in their classifications and outperformed novices, who struggled discriminating phases. The performance of the AI models is comparable to the surgeons in the survey, with improvements when temporal context was incorporated in both cases. Conclusion: SPR is an inherently complex task for expert surgeons and computer vision, where both perform equally well when given the same context. Performance increases when temporal information is provided. Surgical tools and organs form the key landmarks for human interpretation and are expected to shape the future of automated SPR.

Surg-3M: A Dataset and Foundation Model for Perception in Surgical Settings

Advancements in computer-assisted surgical procedures heavily rely on accurate visual data interpretation from camera systems used during surgeries. Traditional open-access datasets focusing on surgical procedures are often limited by their small size, typically consisting of fewer than 100 videos with less than 100K images. To address these constraints, a new dataset called Surg-3M has been compiled using a novel aggregation pipeline that collects high-resolution videos from online sources. Featuring an extensive collection of over 4K surgical videos and more than 3 million high-quality images from multiple procedure types, Surg-3M offers a comprehensive resource surpassing existing alternatives in size and scope, including two novel tasks. To demonstrate the effectiveness of this dataset, we present SurgFM, a self-supervised foundation model pretrained on Surg-3M that achieves impressive results in downstream tasks such as surgical phase recognition, action recognition, and tool presence detection. Combining key components from ConvNeXt, DINO, and an innovative augmented distillation method, SurgFM exhibits exceptional performance compared to specialist architectures across various benchmarks. Our experimental results show that SurgFM outperforms state-of-the-art models in multiple downstream tasks, including significant gains in surgical phase recognition (+8.9pp, +4.7pp, and +3.9pp of Jaccard in AutoLaparo, M2CAI16, and Cholec80), action recognition (+3.1pp of mAP in CholecT50) and tool presence detection (+4.6pp of mAP in Cholec80). Moreover, even when using only half of the data, SurgFM outperforms state-of-the-art models in AutoLaparo and achieves state-of-the-art performance in Cholec80. Both Surg-3M and SurgFM have significant potential to accelerate progress towards developing autonomous robotic surgery systems.

UltraFlwr -- An Efficient Federated Medical and Surgical Object Detection Framework

Object detection shows promise for medical and surgical applications such as cell counting and tool tracking. However, its faces multiple real-world edge deployment challenges including limited high-quality annotated data, data sharing restrictions, and computational constraints. In this work, we introduce UltraFlwr, a framework for federated medical and surgical object detection. By leveraging Federated Learning (FL), UltraFlwr enables decentralized model training across multiple sites without sharing raw data. To further enhance UltraFlwr's efficiency, we propose YOLO-PA, a set of novel Partial Aggregation (PA) strategies specifically designed for YOLO models in FL. YOLO-PA significantly reduces communication overhead by up to 83% per round while maintaining performance comparable to Full Aggregation (FA) strategies. Our extensive experiments on BCCD and m2cai16-tool-locations datasets demonstrate that YOLO-PA not only provides better client models compared to client-wise centralized training and FA strategies, but also facilitates efficient training and deployment across resource-constrained edge devices. Further, we also establish one of the first benchmarks in federated medical and surgical object detection. This paper advances the feasibility of training and deploying detection models on the edge, making federated object detection more practical for time-critical and resource-constrained medical and surgical applications. UltraFlwr is publicly available at https://github.com/KCL-BMEIS/UltraFlwr.

OOD-SEG: Out-Of-Distribution detection for image SEGmentation with sparse multi-class positive-only annotations

Despite significant advancements, segmentation based on deep neural networks in medical and surgical imaging faces several challenges, two of which we aim to address in this work. First, acquiring complete pixel-level segmentation labels for medical images is time-consuming and requires domain expertise. Second, typical segmentation pipelines cannot detect out-of-distribution (OOD) pixels, leaving them prone to spurious outputs during deployment. In this work, we propose a novel segmentation approach exploiting OOD detection that learns only from sparsely annotated pixels from multiple positive-only classes. %but \emph{no background class} annotation. These multi-class positive annotations naturally fall within the in-distribution (ID) set. Unlabelled pixels may contain positive classes but also negative ones, including what is typically referred to as \emph{background} in standard segmentation formulations. Here, we forgo the need for background annotation and consider these together with any other unseen classes as part of the OOD set. Our framework can integrate, at a pixel-level, any OOD detection approaches designed for classification tasks. To address the lack of existing OOD datasets and established evaluation metric for medical image segmentation, we propose a cross-validation strategy that treats held-out labelled classes as OOD. Extensive experiments on both multi-class hyperspectral and RGB surgical imaging datasets demonstrate the robustness and generalisation capability of our proposed framework.

Scribble-Based Interactive Segmentation of Medical Hyperspectral Images

Hyperspectral imaging (HSI) is an advanced medical imaging modality that captures optical data across a broad spectral range, providing novel insights into the biochemical composition of tissues. HSI may enable precise differentiation between various tissue types and pathologies, making it particularly valuable for tumour detection, tissue classification, and disease diagnosis. Deep learning-based segmentation methods have shown considerable advancements, offering automated and accurate results. However, these methods face challenges with HSI datasets due to limited annotated data and discrepancies from hardware and acquisition techniques~\cite{clancy2020surgical,studier2023heiporspectral}. Variability in clinical protocols also leads to different definitions of structure boundaries. Interactive segmentation methods, utilizing user knowledge and clinical insights, can overcome these issues and achieve precise segmentation results \cite{zhao2013overview}. This work introduces a scribble-based interactive segmentation framework for medical hyperspectral images. The proposed method utilizes deep learning for feature extraction and a geodesic distance map generated from user-provided scribbles to obtain the segmentation results. The experiment results show that utilising the geodesic distance maps based on deep learning-extracted features achieved better segmentation results than geodesic distance maps directly generated from hyperspectral images, reconstructed RGB images, or Euclidean distance maps.

Vision and Contact based Optimal Control for Autonomous Trocar Docking

Future operating theatres will be equipped with robots to perform various surgical tasks including, for example, endoscope control. Human-in-the-loop supervisory control architectures where the surgeon selects from several autonomous sequences is already being successfully applied in preclinical tests. Inserting an endoscope into a trocar or introducer is a key step for every keyhole surgical procedure -- hereafter we will only refer to this device as a "trocar". Our goal is to develop a controller for autonomous trocar docking. Autonomous trocar docking is a version of the peg-in-hole problem. Extensive work in the robotics literature addresses this problem. The peg-in-hole problem has been widely studied in the context of assembly where, typically, the hole is considered static and rigid to interaction. In our case, however, the trocar is not fixed and responds to interaction. We consider a variety of surgical procedures where surgeons will utilize contact between the endoscope and trocar in order to complete the insertion successfully. To the best of our knowledge, we have not found literature that explores this particular generalization of the problem directly. Our primary contribution in this work is an optimal control formulation for automated trocar docking. We use a nonlinear optimization program to model the task, minimizing a cost function subject to constraints to find optimal joint configurations. The controller incorporates a geometric model for insertion and a force-feedback (FF) term to ensure patient safety by preventing excessive interaction forces with the trocar. Experiments, demonstrated on a real hardware lab setup, validate the approach. Our method successfully achieves trocar insertion on our real robot lab setup, and simulation trials demonstrate its ability to reduce interaction forces.

A self-supervised and adversarial approach to hyperspectral demosaicking and RGB reconstruction in surgical imaging

Hyperspectral imaging holds promises in surgical imaging by offering biological tissue differentiation capabilities with detailed information that is invisible to the naked eye. For intra-operative guidance, real-time spectral data capture and display is mandated. Snapshot mosaic hyperspectral cameras are currently seen as the most suitable technology given this requirement. However, snapshot mosaic imaging requires a demosaicking algorithm to fully restore the spatial and spectral details in the images. Modern demosaicking approaches typically rely on synthetic datasets to develop supervised learning methods, as it is practically impossible to simultaneously capture both snapshot and high-resolution spectral images of the exact same surgical scene. In this work, we present a self-supervised demosaicking and RGB reconstruction method that does not depend on paired high-resolution data as ground truth. We leverage unpaired standard high-resolution surgical microscopy images, which only provide RGB data but can be collected during routine surgeries. Adversarial learning complemented by self-supervised approaches are used to drive our hyperspectral-based RGB reconstruction into resembling surgical microscopy images and increasing the spatial resolution of our demosaicking. The spatial and spectral fidelity of the reconstructed hyperspectral images have been evaluated quantitatively. Moreover, a user study was conducted to evaluate the RGB visualisation generated from these spectral images. Both spatial detail and colour accuracy were assessed by neurosurgical experts. Our proposed self-supervised demosaicking method demonstrates improved results compared to existing methods, demonstrating its potential for seamless integration into intra-operative workflows.

Nonrigid Reconstruction of Freehand Ultrasound without a Tracker

Reconstructing 2D freehand Ultrasound (US) frames into 3D space without using a tracker has recently seen advances with deep learning. Predicting good frame-to-frame rigid transformations is often accepted as the learning objective, especially when the ground-truth labels from spatial tracking devices are inherently rigid transformations. Motivated by a) the observed nonrigid deformation due to soft tissue motion during scanning, and b) the highly sensitive prediction of rigid transformation, this study investigates the methods and their benefits in predicting nonrigid transformations for reconstructing 3D US. We propose a novel co-optimisation algorithm for simultaneously estimating rigid transformations among US frames, supervised by ground-truth from a tracker, and a nonrigid deformation, optimised by a regularised registration network. We show that these two objectives can be either optimised using meta-learning or combined by weighting. A fast scattered data interpolation is also developed for enabling frequent reconstruction and registration of non-parallel US frames, during training. With a new data set containing over 357,000 frames in 720 scans, acquired from 60 subjects, the experiments demonstrate that, due to an expanded thus easier-to-optimise solution space, the generalisation is improved with the added deformation estimation, with respect to the rigid ground-truth. The global pixel reconstruction error (assessing accumulative prediction) is lowered from 18.48 to 16.51 mm, compared with baseline rigid-transformation-predicting methods. Using manually identified landmarks, the proposed co-optimisation also shows potentials in compensating nonrigid tissue motion at inference, which is not measurable by tracker-provided ground-truth. The code and data used in this paper are made publicly available at https://github.com/QiLi111/NR-Rec-FUS.