Stefanie Speidel

Dominik Rivoir, Isabel Funke, Stefanie Speidel

Batch Normalization's (BN) unique property of depending on other samples in a batch is known to cause problems in several tasks, including sequence modeling. Yet, BN-related issues are hardly studied for long video understanding, despite the ubiquitous use of BN in CNNs (Convolutional Neural Networks) for feature extraction. Especially in surgical workflow analysis, where the lack of pretrained feature extractors has led to complex, multi-stage training pipelines, limited awareness of BN issues may have hidden the benefits of training CNNs and temporal models end to end. In this paper, we analyze pitfalls of BN in video learning, including issues specific to online tasks such as a 'cheating' effect in anticipation. We observe that BN's properties create major obstacles for end-to-end learning. However, using BN-free backbones, even simple CNN-LSTMs beat the state of the art {\color{\colorrevtwo}on three surgical workflow benchmarks} by utilizing adequate end-to-end training strategies which maximize temporal context. We conclude that awareness of BN's pitfalls is crucial for effective end-to-end learning in surgical tasks. By reproducing results on natural-video datasets, we hope our insights will benefit other areas of video learning as well. Code is available at: \url{https://gitlab.com/nct_tso_public/pitfalls_bn}

Peng Liu, Jonas Golde, Joseph Morgenstern et al.

Purpose: Middle ear infection is the most prevalent inflammatory disease, especially among the pediatric population. Current diagnostic methods are subjective and depend on visual cues from an otoscope, which is limited for otologists to identify pathology. To address this shortcoming, endoscopic optical coherence tomography (OCT) provides both morphological and functional in-vivo measurements of the middle ear. However, due to the shadow of prior structures, interpretation of OCT images is challenging and time-consuming. To facilitate fast diagnosis and measurement, improvement in the readability of OCT data is achieved by merging morphological knowledge from ex-vivo middle ear models with OCT volumetric data, so that OCT applications can be further promoted in daily clinical settings. Methods: We propose C2P-Net: a two-staged non-rigid registration pipeline for complete to partial point clouds, which are sampled from ex-vivo and in-vivo OCT models, respectively. To overcome the lack of labeled training data, a fast and effective generation pipeline in Blender3D is designed to simulate middle ear shapes and extract in-vivo noisy and partial point clouds. Results: We evaluate the performance of C2P-Net through experiments on both synthetic and real OCT datasets. The results demonstrate that C2P-Net is generalized to unseen middle ear point clouds and capable of handling realistic noise and incompleteness in synthetic and real OCT data. Conclusion: In this work, we aim to enable diagnosis of middle ear structures with the assistance of OCT images. We propose C2P-Net: a two-staged non-rigid registration pipeline for point clouds to support the interpretation of in-vivo noisy and partial OCT images for the first time. Code is available at: https://gitlab.com/nct\_tso\_public/c2p-net.

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.

Isabel Funke, Dominik Rivoir, Stefanie Krell et al.

Objective: To enable context-aware computer assistance in the operating room of the future, cognitive systems need to understand automatically which surgical phase is being performed by the medical team. The primary source of information for surgical phase recognition is typically video, which presents two challenges: extracting meaningful features from the video stream and effectively modeling temporal information in the sequence of visual features. Methods: For temporal modeling, attention mechanisms have gained popularity due to their ability to capture long-range dependencies. In this paper, we explore design choices for attention in existing temporal models for surgical phase recognition and propose a novel approach that uses attention more effectively and does not require hand-crafted constraints: TUNeS, an efficient and simple temporal model that incorporates self-attention at the core of a convolutional U-Net structure. In addition, we propose to train the feature extractor, a standard CNN, together with an LSTM on preferably long video segments, i.e., with long temporal context. Results: In our experiments, almost all temporal models performed better on top of feature extractors that were trained with longer temporal context. On these contextualized features, TUNeS achieves state-of-the-art results on the Cholec80 dataset. Conclusion: This study offers new insights on how to use attention mechanisms to build accurate and efficient temporal models for surgical phase recognition. Significance: Implementing automatic surgical phase recognition is essential to automate the analysis and optimization of surgical workflows and to enable context-aware computer assistance during surgery, thus ultimately improving patient care.

Aneeq Zia, Kiran Bhattacharyya, Xi Liu et al.

Timely and effective feedback within surgical training plays a critical role in developing the skills required to perform safe and efficient surgery. Feedback from expert surgeons, while especially valuable in this regard, is challenging to acquire due to their typically busy schedules, and may be subject to biases. Formal assessment procedures like OSATS and GEARS attempt to provide objective measures of skill, but remain time-consuming. With advances in machine learning there is an opportunity for fast and objective automated feedback on technical skills. The SimSurgSkill 2021 challenge (hosted as a sub-challenge of EndoVis at MICCAI 2021) aimed to promote and foster work in this endeavor. Using virtual reality (VR) surgical tasks, competitors were tasked with localizing instruments and predicting surgical skill. Here we summarize the winning approaches and how they performed. Using this publicly available dataset and results as a springboard, future work may enable more efficient training of surgeons with advances in surgical data science. The dataset can be accessed from https://console.cloud.google.com/storage/browser/isi-simsurgskill-2021.

Nithya Bhasker, Stefan Leger, Alexander Zwanenburg et al.

Graph neural networks (GNNs) are becoming increasingly popular in the medical domain for the tasks of disease classification and outcome prediction. Since patient data is not readily available as a graph, most existing methods either manually define a patient graph, or learn a latent graph based on pairwise similarities between the patients. There are also hypergraph neural network (HGNN)-based methods that were introduced recently to exploit potential higher order associations between the patients by representing them as a hypergraph. In this work, we propose a patient hypergraph network (PHGN), which has been investigated in an inductive learning setup for binary outcome prediction in oropharyngeal cancer (OPC) patients using computed tomography (CT)-based radiomic features for the first time. Additionally, the proposed model was extended to perform time-to-event analyses, and compared with GNN and baseline linear models.

Danush Kumar Venkatesh, Isabel Funke, Micha Pfeiffer et al.

Computer-assisted interventions can improve intra-operative guidance, particularly through deep learning methods that harness the spatiotemporal information in surgical videos. However, the severe data imbalance often found in surgical video datasets hinders the development of high-performing models. In this work, we aim to overcome the data imbalance by synthesizing surgical videos. We propose a unique two-stage, text-conditioned diffusion-based method to generate high-fidelity surgical videos for under-represented classes. Our approach conditions the generation process on text prompts and decouples spatial and temporal modeling by utilizing a 2D latent diffusion model to capture spatial content and then integrating temporal attention layers to ensure temporal consistency. Furthermore, we introduce a rejection sampling strategy to select the most suitable synthetic samples, effectively augmenting existing datasets to address class imbalance. We evaluate our method on two downstream tasks-surgical action recognition and intra-operative event prediction-demonstrating that incorporating synthetic videos from our approach substantially enhances model performance. We open-source our implementation at https://gitlab.com/nct_tso_public/surgvgen.

Micha Pfeiffer, Hannes Kenngott, Anas Preukschas et al.

Purpose: The data which is available to surgeons before, during and after surgery is steadily increasing in quantity as well as diversity. When planning a patient's treatment, this large amount of information can be difficult to interpret. To aid in processing the information, new methods need to be found to present multi-modal patient data, ideally combining textual, imagery, temporal and 3D data in a holistic and context-aware system. Methods: We present an open-source framework which allows handling of patient data in a virtual reality (VR) environment. By using VR technology, the workspace available to the surgeon is maximized and 3D patient data is rendered in stereo, which increases depth perception. The framework organizes the data into workspaces and contains tools which allow users to control, manipulate and enhance the data. Due to the framework's modular design, it can easily be adapted and extended for various clinical applications. Results: The framework was evaluated by clinical personnel (77 participants). The majority of the group stated that a complex surgical situation is easier to comprehend by using the framework, and that it is very well suited for education. Furthermore, the application to various clinical scenarios - including the simulation of excitation-propagation in the human atrium - demonstrated the framework's adaptability. As a feasibility study, the framework was used during the planning phase of the surgical removal of a large central carcinoma from a patient's liver. Conclusion: The clinical evaluation showed a large potential and high acceptance for the VR environment in a medical context. The various applications confirmed that the framework is easily extended and can be used in real-time simulation as well as for the manipulation of complex anatomical structures.

Alexander C. Jenke, Sebastian Bodenstedt, Fiona R. Kolbinger et al.

Understanding a surgical scene is crucial for computer-assisted surgery systems to provide any intelligent assistance functionality. One way of achieving this scene understanding is via scene segmentation, where every pixel of a frame is classified and therefore identifies the visible structures and tissues. Progress on fully segmenting surgical scenes has been made using machine learning. However, such models require large amounts of annotated training data, containing examples of all relevant object classes. Such fully annotated datasets are hard to create, as every pixel in a frame needs to be annotated by medical experts and, therefore, are rarely available. In this work, we propose a method to combine multiple partially annotated datasets, which provide complementary annotations, into one model, enabling better scene segmentation and the use of multiple readily available datasets. Our method aims to combine available data with complementary labels by leveraging mutual exclusive properties to maximize information. Specifically, we propose to use positive annotations of other classes as negative samples and to exclude background pixels of binary annotations, as we cannot tell if they contain a class not annotated but predicted by the model. We evaluate our method by training a DeepLabV3 on the publicly available Dresden Surgical Anatomy Dataset, which provides multiple subsets of binary segmented anatomical structures. Our approach successfully combines 6 classes into one model, increasing the overall Dice Score by 4.4% compared to an ensemble of models trained on the classes individually. By including information on multiple classes, we were able to reduce confusion between stomach and colon by 24%. Our results demonstrate the feasibility of training a model on multiple datasets. This paves the way for future work further alleviating the need for one large, fully segmented datasets.

Max Kirchner, Alexander C. Jenke, Sebastian Bodenstedt et al.

Purpose: In this study, we investigate the training of foundation models using federated learning to address data-sharing limitations and enable collaborative model training without data transfer for minimally invasive surgery. Methods: Inspired by the EndoViT study, we adapt the Masked Autoencoder for federated learning, enhancing it with adaptive Sharpness-Aware Minimization (FedSAM) and Stochastic Weight Averaging (SWA). Our model is pretrained on the Endo700k dataset collection and later fine-tuned and evaluated for tasks such as Semantic Segmentation, Action Triplet Recognition, and Surgical Phase Recognition. Results: Our findings demonstrate that integrating adaptive FedSAM into the federated MAE approach improves pretraining, leading to a reduction in reconstruction loss per patch. The application of FL-EndoViT in surgical downstream tasks results in performance comparable to CEN-EndoViT. Furthermore, FL-EndoViT exhibits advantages over CEN-EndoViT in surgical scene segmentation when data is limited and in action triplet recognition when large datasets are used. Conclusion: These findings highlight the potential of federated learning for privacy-preserving training of surgical foundation models, offering a robust and generalizable solution for surgical data science. Effective collaboration requires adapting federated learning methods, such as the integration of FedSAM, which can accommodate the inherent data heterogeneity across institutions. In future, exploring FL in video-based models may enhance these capabilities by incorporating spatiotemporal dynamics crucial for real-world surgical environments.

Max Kirchner, Hanna Hoffmann, Alexander C. Jenke et al.

Purpose: The FedSurg challenge was designed to benchmark the state of the art in federated learning for surgical video classification. Its goal was to assess how well current methods generalize to unseen clinical centers and adapt through local fine-tuning while enabling collaborative model development without sharing patient data. Methods: Participants developed strategies to classify inflammation stages in appendicitis using a preliminary version of the multi-center Appendix300 video dataset. The challenge evaluated two tasks: generalization to an unseen center and center-specific adaptation after fine-tuning. Submitted approaches included foundation models with linear probing, metric learning with triplet loss, and various FL aggregation schemes (FedAvg, FedMedian, FedSAM). Performance was assessed using F1-score and Expected Cost, with ranking robustness evaluated via bootstrapping and statistical testing. Results: In the generalization task, performance across centers was limited. In the adaptation task, all teams improved after fine-tuning, though ranking stability was low. The ViViT-based submission achieved the strongest overall performance. The challenge highlighted limitations in generalization, sensitivity to class imbalance, and difficulties in hyperparameter tuning in decentralized training, while spatiotemporal modeling and context-aware preprocessing emerged as promising strategies. Conclusion: The FedSurg Challenge establishes the first benchmark for evaluating FL strategies in surgical video classification. Findings highlight the trade-off between local personalization and global robustness, and underscore the importance of architecture choice, preprocessing, and loss design. This benchmarking offers a reference point for future development of imbalance-aware, adaptive, and robust FL methods in clinical surgical AI.

Karin A. Olthof, Matteo Fusagli, Bianca Güttner et al.

Background: The aim of this study was to develop and evaluate a deep learning-based automated segmentation method for hepatic anatomy (i.e., parenchyma, tumors, portal vein, hepatic vein and biliary tree) from the hepatobiliary phase of gadoxetic acid-enhanced MRI. This method should ease the clinical workflow of preoperative planning. Methods: Manual segmentation was performed on hepatobiliary phase MRI scans from 90 consecutive patients who underwent liver surgery between January 2020 and October 2023. A deep learning network (nnU-Net v1) was trained on 72 patients with an extra focus on thin structures and topography preservation. Performance was evaluated on an 18-patient test set by comparing automated and manual segmentations using Dice similarity coefficient (DSC). Following clinical integration, 10 segmentations (assessment dataset) were generated using the network and manually refined for clinical use to quantify required adjustments using DSC. Results: In the test set, DSCs were 0.97+/-0.01 for liver parenchyma, 0.80+/-0.04 for hepatic vein, 0.79+/-0.07 for biliary tree, 0.77+/-0.17 for tumors, and 0.74+/-0.06 for portal vein. Average tumor detection rate was 76.6+/-24.1%, with a median of one false-positive per patient. The assessment dataset showed minor adjustments were required for clinical use of the 3D models, with high DSCs for parenchyma (1.00+/-0.00), portal vein (0.98+/-0.01) and hepatic vein (0.95+/-0.07). Tumor segmentation exhibited greater variability (DSC 0.80+/-0.27). During prospective clinical use, the model detected three additional tumors initially missed by radiologists. Conclusions: The proposed nnU-Net-based segmentation method enables accurate and automated delineation of hepatic anatomy. This enables 3D planning to be applied efficiently as a standard-of-care for every patient undergoing liver surgery.

Zhongyu Chen, Chenyang Li, Xuesong Li et al.

Precise needle alignment is essential for percutaneous needle insertion in robotic ultrasound-guided procedures. However, inherent challenges such as speckle noise, needle-like artifacts, and low image resolution make robust needle detection difficult, particularly when visibility is reduced or lost. In this paper, we propose a method to restore needle alignment when the ultrasound imaging plane and the needle insertion plane are misaligned. Unlike many existing approaches that rely heavily on needle visibility in ultrasound images, our method uses a more robust feature by periodically vibrating the needle using a mechanical system. Specifically, we propose a vibration-based energy metric that remains effective even when the needle is fully out of plane. Using this metric, we develop a control strategy to reposition the ultrasound probe in response to misalignments between the imaging plane and the needle insertion plane in both translation and rotation. Experiments conducted on ex-vivo porcine tissue samples using a dual-arm robotic ultrasound-guided needle insertion system demonstrate the effectiveness of the proposed approach. The experimental results show the translational error of 0.41$\pm$0.27 mm and the rotational error of 0.51$\pm$0.19 degrees.

Tobias Rueckert, David Rauber, Raphaela Maerkl et al.

Reliable recognition and localization of surgical instruments in endoscopic video recordings are foundational for a wide range of applications in computer- and robot-assisted minimally invasive surgery (RAMIS), including surgical training, skill assessment, and autonomous assistance. However, robust performance under real-world conditions remains a significant challenge. Incorporating surgical context - such as the current procedural phase - has emerged as a promising strategy to improve robustness and interpretability. To address these challenges, we organized the Surgical Procedure Phase, Keypoint, and Instrument Recognition (PhaKIR) sub-challenge as part of the Endoscopic Vision (EndoVis) challenge at MICCAI 2024. We introduced a novel, multi-center dataset comprising thirteen full-length laparoscopic cholecystectomy videos collected from three distinct medical institutions, with unified annotations for three interrelated tasks: surgical phase recognition, instrument keypoint estimation, and instrument instance segmentation. Unlike existing datasets, ours enables joint investigation of instrument localization and procedural context within the same data while supporting the integration of temporal information across entire procedures. We report results and findings in accordance with the BIAS guidelines for biomedical image analysis challenges. The PhaKIR sub-challenge advances the field by providing a unique benchmark for developing temporally aware, context-driven methods in RAMIS and offers a high-quality resource to support future research in surgical scene understanding.

Isabel Funke, Dominik Rivoir, Stefanie Speidel

Surgical phase recognition is a basic component for different context-aware applications in computer- and robot-assisted surgery. In recent years, several methods for automatic surgical phase recognition have been proposed, showing promising results. However, a meaningful comparison of these methods is difficult due to differences in the evaluation process and incomplete reporting of evaluation details. In particular, the details of metric computation can vary widely between different studies. To raise awareness of potential inconsistencies, this paper summarizes common deviations in the evaluation of phase recognition algorithms on the Cholec80 benchmark. In addition, a structured overview of previously reported evaluation results on Cholec80 is provided, taking known differences in evaluation protocols into account. Greater attention to evaluation details could help achieve more consistent and comparable results on the surgical phase recognition task, leading to more reliable conclusions about advancements in the field and, finally, translation into clinical practice.

Aneeq Zia, Max Berniker, Rogerio Garcia Nespolo et al.

Robotic assisted (RA) surgery promises to transform surgical intervention. Intuitive Surgical is committed to fostering these changes and the machine learning models and algorithms that will enable them. With these goals in mind we have invited the surgical data science community to participate in a yearly competition hosted through the Medical Imaging Computing and Computer Assisted Interventions (MICCAI) conference. With varying changes from year to year, we have challenged the community to solve difficult machine learning problems in the context of advanced RA applications. Here we document the results of these challenges, focusing on surgical tool localization (SurgToolLoc). The publicly released dataset that accompanies these challenges is detailed in a separate paper arXiv:2501.09209 [1].

Martin Wagner, Beat-Peter Müller-Stich, Anna Kisilenko et al.

PURPOSE: Surgical workflow and skill analysis are key technologies for the next generation of cognitive surgical assistance systems. These systems could increase the safety of the operation through context-sensitive warnings and semi-autonomous robotic assistance or improve training of surgeons via data-driven feedback. In surgical workflow analysis up to 91% average precision has been reported for phase recognition on an open data single-center dataset. In this work we investigated the generalizability of phase recognition algorithms in a multi-center setting including more difficult recognition tasks such as surgical action and surgical skill. METHODS: To achieve this goal, a dataset with 33 laparoscopic cholecystectomy videos from three surgical centers with a total operation time of 22 hours was created. Labels included annotation of seven surgical phases with 250 phase transitions, 5514 occurences of four surgical actions, 6980 occurences of 21 surgical instruments from seven instrument categories and 495 skill classifications in five skill dimensions. The dataset was used in the 2019 Endoscopic Vision challenge, sub-challenge for surgical workflow and skill analysis. Here, 12 teams submitted their machine learning algorithms for recognition of phase, action, instrument and/or skill assessment. RESULTS: F1-scores were achieved for phase recognition between 23.9% and 67.7% (n=9 teams), for instrument presence detection between 38.5% and 63.8% (n=8 teams), but for action recognition only between 21.8% and 23.3% (n=5 teams). The average absolute error for skill assessment was 0.78 (n=1 team). CONCLUSION: Surgical workflow and skill analysis are promising technologies to support the surgical team, but are not solved yet, as shown by our comparison of algorithms. This novel benchmark can be used for comparable evaluation and validation of future work.

Aneeq Zia, Kiran Bhattacharyya, Xi Liu et al.

Surgical data science is revolutionizing minimally invasive surgery by enabling context-aware applications. However, many challenges exist around surgical data (and health data, more generally) needed to develop context-aware models. This work - presented as part of the Endoscopic Vision (EndoVis) challenge at the Medical Image Computing and Computer Assisted Intervention (MICCAI) 2020 conference - seeks to explore the potential for visual domain adaptation in surgery to overcome data privacy concerns. In particular, we propose to use video from virtual reality (VR) simulations of surgical exercises in robotic-assisted surgery to develop algorithms to recognize tasks in a clinical-like setting. We present the performance of the different approaches to solve visual domain adaptation developed by challenge participants. Our analysis shows that the presented models were unable to learn meaningful motion based features form VR data alone, but did significantly better when small amount of clinical-like data was also made available. Based on these results, we discuss promising methods and further work to address the problem of visual domain adaptation in surgical data science. We also release the challenge dataset publicly at https://www.synapse.org/surgvisdom2020.

P. J. "Eddie'' Edwards, Dimitris Psychogyios, Stefanie Speidel et al.

In computer vision, reference datasets have been highly successful in promoting algorithmic development in stereo reconstruction. Surgical scenes gives rise to specific problems, including the lack of clear corner features, highly specular surfaces and the presence of blood and smoke. Publicly available datasets have been produced using CT and either phantom images or biological tissue samples covering a relatively small region of the endoscope field-of-view. We present a stereo-endoscopic reconstruction validation dataset based on CT (SERV-CT). Two {\it ex vivo} small porcine full torso cadavers were placed within the view of the endoscope with both the endoscope and target anatomy visible in the CT scan. Orientation of the endoscope was manually aligned to the stereoscopic view. Reference disparities and occlusions were calculated for 8 stereo pairs from each sample. For the second sample an RGB surface was acquired to aid alignment of smooth, featureless surfaces. Repeated manual alignments showed an RMS disparity accuracy of ~2 pixels and a depth accuracy of ~2mm. The reference dataset includes endoscope image pairs with corresponding calibration, disparities, depths and occlusions covering the majority of the endoscopic image and a range of tissue types. Smooth specular surfaces and images with significant variation of depth are included. We assessed the performance of various stereo algorithms from online available repositories. There is a significant variation between algorithms, highlighting some of the challenges of surgical endoscopic images. The SERV-CT dataset provides an easy to use stereoscopic validation for surgical applications with smooth reference disparities and depths with coverage over the majority of the endoscopic images. This complements existing resources well and we hope will aid the development of surgical endoscopic anatomical reconstruction algorithms.

Lena Maier-Hein, Matthias Eisenmann, Duygu Sarikaya et al.

Recent developments in data science in general and machine learning in particular have transformed the way experts envision the future of surgery. Surgical Data Science (SDS) is a new research field that aims to improve the quality of interventional healthcare through the capture, organization, analysis and modeling of data. While an increasing number of data-driven approaches and clinical applications have been studied in the fields of radiological and clinical data science, translational success stories are still lacking in surgery. In this publication, we shed light on the underlying reasons and provide a roadmap for future advances in the field. Based on an international workshop involving leading researchers in the field of SDS, we review current practice, key achievements and initiatives as well as available standards and tools for a number of topics relevant to the field, namely (1) infrastructure for data acquisition, storage and access in the presence of regulatory constraints, (2) data annotation and sharing and (3) data analytics. We further complement this technical perspective with (4) a review of currently available SDS products and the translational progress from academia and (5) a roadmap for faster clinical translation and exploitation of the full potential of SDS, based on an international multi-round Delphi process.

Dominik Rivoir, Sebastian Bodenstedt, Isabel Funke et al.

Intra-operative anticipation of instrument usage is a necessary component for context-aware assistance in surgery, e.g. for instrument preparation or semi-automation of robotic tasks. However, the sparsity of instrument occurrences in long videos poses a challenge. Current approaches are limited as they assume knowledge on the timing of future actions or require dense temporal segmentations during training and inference. We propose a novel learning task for anticipation of instrument usage in laparoscopic videos that overcomes these limitations. During training, only sparse instrument annotations are required and inference is done solely on image data. We train a probabilistic model to address the uncertainty associated with future events. Our approach outperforms several baselines and is competitive to a variant using richer annotations. We demonstrate the model's ability to quantify task-relevant uncertainties. To the best of our knowledge, we are the first to propose a method for anticipating instruments in surgery.

Micha Pfeiffer, Carina Riediger, Stefan Leger et al.

Non-rigid registration is a key component in soft-tissue navigation. We focus on laparoscopic liver surgery, where we register the organ model obtained from a preoperative CT scan to the intraoperative partial organ surface, reconstructed from the laparoscopic video. This is a challenging task due to sparse and noisy intraoperative data, real-time requirements and many unknowns - such as tissue properties and boundary conditions. Furthermore, establishing correspondences between pre- and intraoperative data can be extremely difficult since the liver usually lacks distinct surface features and the used imaging modalities suffer from very different types of noise. In this work, we train a convolutional neural network to perform both the search for surface correspondences as well as the non-rigid registration in one step. The network is trained on physically accurate biomechanical simulations of randomly generated, deforming organ-like structures. This enables the network to immediately generalize to a new patient organ without the need to re-train. We add various amounts of noise to the intraoperative surfaces during training, making the network robust to noisy intraoperative data. During inference, the network outputs the displacement field which matches the preoperative volume to the partial intraoperative surface. In multiple experiments, we show that the network translates well to real data while maintaining a high inference speed. Our code is made available online.

Lena Maier-Hein, Martin Wagner, Tobias Ross et al.

Image-based tracking of medical instruments is an integral part of surgical data science applications. Previous research has addressed the tasks of detecting, segmenting and tracking medical instruments based on laparoscopic video data. However, the proposed methods still tend to fail when applied to challenging images and do not generalize well to data they have not been trained on. This paper introduces the Heidelberg Colorectal (HeiCo) data set - the first publicly available data set enabling comprehensive benchmarking of medical instrument detection and segmentation algorithms with a specific emphasis on method robustness and generalization capabilities. Our data set comprises 30 laparoscopic videos and corresponding sensor data from medical devices in the operating room for three different types of laparoscopic surgery. Annotations include surgical phase labels for all video frames as well as information on instrument presence and corresponding instance-wise segmentation masks for surgical instruments (if any) in more than 10,000 individual frames. The data has successfully been used to organize international competitions within the Endoscopic Vision Challenges 2017 and 2019.

Tobias Ross, Annika Reinke, Peter M. Full et al.

Intraoperative tracking of laparoscopic instruments is often a prerequisite for computer and robotic-assisted interventions. While numerous methods for detecting, segmenting and tracking of medical instruments based on endoscopic video images have been proposed in the literature, key limitations remain to be addressed: Firstly, robustness, that is, the reliable performance of state-of-the-art methods when run on challenging images (e.g. in the presence of blood, smoke or motion artifacts). Secondly, generalization; algorithms trained for a specific intervention in a specific hospital should generalize to other interventions or institutions. In an effort to promote solutions for these limitations, we organized the Robust Medical Instrument Segmentation (ROBUST-MIS) challenge as an international benchmarking competition with a specific focus on the robustness and generalization capabilities of algorithms. For the first time in the field of endoscopic image processing, our challenge included a task on binary segmentation and also addressed multi-instance detection and segmentation. The challenge was based on a surgical data set comprising 10,040 annotated images acquired from a total of 30 surgical procedures from three different types of surgery. The validation of the competing methods for the three tasks (binary segmentation, multi-instance detection and multi-instance segmentation) was performed in three different stages with an increasing domain gap between the training and the test data. The results confirm the initial hypothesis, namely that algorithm performance degrades with an increasing domain gap. While the average detection and segmentation quality of the best-performing algorithms is high, future research should concentrate on detection and segmentation of small, crossing, moving and transparent instrument(s) (parts).

Dominik Rivoir, Sebastian Bodenstedt, Felix von Bechtolsheim et al.

Estimating the remaining surgery duration (RSD) during surgical procedures can be useful for OR planning and anesthesia dose estimation. With the recent success of deep learning-based methods in computer vision, several neural network approaches have been proposed for fully automatic RSD prediction based solely on visual data from the endoscopic camera. We investigate whether RSD prediction can be improved using unsupervised temporal video segmentation as an auxiliary learning task. As opposed to previous work, which presented supervised surgical phase recognition as auxiliary task, we avoid the need for manual annotations by proposing a similar but unsupervised learning objective which clusters video sequences into temporally coherent segments. In multiple experimental setups, results obtained by learning the auxiliary task are incorporated into a deep RSD model through feature extraction, pretraining or regularization. Further, we propose a novel loss function for RSD training which attempts to counteract unfavorable characteristics of the RSD ground truth. Using our unsupervised method as an auxiliary task for RSD training, we outperform other self-supervised methods and are comparable to the supervised state-of-the-art. Combined with the novel RSD loss, we slightly outperform the supervised approach.

Isabel Funke, Sebastian Bodenstedt, Florian Oehme et al.

Automatically recognizing surgical gestures is a crucial step towards a thorough understanding of surgical skill. Possible areas of application include automatic skill assessment, intra-operative monitoring of critical surgical steps, and semi-automation of surgical tasks. Solutions that rely only on the laparoscopic video and do not require additional sensor hardware are especially attractive as they can be implemented at low cost in many scenarios. However, surgical gesture recognition based only on video is a challenging problem that requires effective means to extract both visual and temporal information from the video. Previous approaches mainly rely on frame-wise feature extractors, either handcrafted or learned, which fail to capture the dynamics in surgical video. To address this issue, we propose to use a 3D Convolutional Neural Network (CNN) to learn spatiotemporal features from consecutive video frames. We evaluate our approach on recordings of robot-assisted suturing on a bench-top model, which are taken from the publicly available JIGSAWS dataset. Our approach achieves high frame-wise surgical gesture recognition accuracies of more than 84%, outperforming comparable models that either extract only spatial features or model spatial and low-level temporal information separately. For the first time, these results demonstrate the benefit of spatiotemporal CNNs for video-based surgical gesture recognition.

Max Allan, Alex Shvets, Thomas Kurmann et al.

In mainstream computer vision and machine learning, public datasets such as ImageNet, COCO and KITTI have helped drive enormous improvements by enabling researchers to understand the strengths and limitations of different algorithms via performance comparison. However, this type of approach has had limited translation to problems in robotic assisted surgery as this field has never established the same level of common datasets and benchmarking methods. In 2015 a sub-challenge was introduced at the EndoVis workshop where a set of robotic images were provided with automatically generated annotations from robot forward kinematics. However, there were issues with this dataset due to the limited background variation, lack of complex motion and inaccuracies in the annotation. In this work we present the results of the 2017 challenge on robotic instrument segmentation which involved 10 teams participating in binary, parts and type based segmentation of articulated da Vinci robotic instruments.

Sebastian Bodenstedt, Martin Wagner, Lars Mündermann et al.

Purpose The course of surgical procedures is often unpredictable, making it difficult to estimate the duration of procedures beforehand. A context-aware method that analyses the workflow of an intervention online and automatically predicts the remaining duration would alleviate these problems. As basis for such an estimate, information regarding the current state of the intervention is required. Methods Today, the operating room contains a diverse range of sensors. During laparoscopic interventions, the endoscopic video stream is an ideal source of such information. Extracting quantitative information from the video is challenging though, due to its high dimensionality. Other surgical devices (e.g. insufflator, lights, etc.) provide data streams which are, in contrast to the video stream, more compact and easier to quantify. Though whether such streams offer sufficient information for estimating the duration of surgery is uncertain. Here, we propose and compare methods, based on convolutional neural networks, for continuously predicting the duration of laparoscopic interventions based on unlabeled data, such as from endoscopic images and surgical device streams. Results The methods are evaluated on 80 laparoscopic interventions of various types, for which surgical device data and the endoscopic video are available. Here the combined method performs best with an overall average error of 37% and an average halftime error of 28%. Conclusion In this paper, we present, to our knowledge, the first approach for online procedure duration prediction using unlabeled endoscopic video data and surgical device data in a laparoscopic setting. We also show that a method incorporating both vision and device data performs better than methods based only on vision, while methods only based on tool usage and surgical device data perform poorly, showing the importance of the visual channel.

Sebastian Bodenstedt, Dominik Rivoir, Alexander Jenke et al.

For many applications in the field of computer assisted surgery, such as providing the position of a tumor, specifying the most probable tool required next by the surgeon or determining the remaining duration of surgery, methods for surgical workflow analysis are a prerequisite. Often machine learning based approaches serve as basis for surgical workflow analysis. In general machine learning algorithms, such as convolutional neural networks (CNN), require large amounts of labeled data. While data is often available in abundance, many tasks in surgical workflow analysis need data annotated by domain experts, making it difficult to obtain a sufficient amount of annotations. The aim of using active learning to train a machine learning model is to reduce the annotation effort. Active learning methods determine which unlabeled data points would provide the most information according to some metric, such as prediction uncertainty. Experts will then be asked to only annotate these data points. The model is then retrained with the new data and used to select further data for annotation. Recently, active learning has been applied to CNN by means of Deep Bayesian Networks (DBN). These networks make it possible to assign uncertainties to predictions. In this paper, we present a DBN-based active learning approach adapted for image-based surgical workflow analysis task. Furthermore, by using a recurrent architecture, we extend this network to video-based surgical workflow analysis. We evaluate these approaches on the Cholec80 dataset by performing instrument presence detection and surgical phase segmentation. Here we are able to show that using a DBN-based active learning approach for selecting what data points to annotate next outperforms a baseline based on randomly selecting data points.

Sebastian Bodenstedt, Antonia Ohnemus, Darko Katic et al.

During laparoscopic surgery, context-aware assistance systems aim to alleviate some of the difficulties the surgeon faces. To ensure that the right information is provided at the right time, the current phase of the intervention has to be known. Real-time locating and classification the surgical tools currently in use are key components of both an activity-based phase recognition and assistance generation. In this paper, we present an image-based approach that detects and classifies tools during laparoscopic interventions in real-time. First, potential instrument bounding boxes are detected using a pixel-wise random forest segmentation. Each of these bounding boxes is then classified using a cascade of random forest. For this, multiple features, such as histograms over hue and saturation, gradients and SURF feature, are extracted from each detected bounding box. We evaluated our approach on five different videos from two different types of procedures. We distinguished between the four most common classes of instruments (LigaSure, atraumatic grasper, aspirator, clip applier) and background. Our method succesfully located up to 86% of all instruments respectively. On manually provided bounding boxes, we achieve a instrument type recognition rate of up to 58% and on automatically detected bounding boxes up to 49%. To our knowledge, this is the first approach that allows an image-based classification of surgical tools in a laparoscopic setting in real-time.

Sebastian Bodenstedt, Max Allan, Anthony Agustinos et al.

Intraoperative segmentation and tracking of minimally invasive instruments is a prerequisite for computer- and robotic-assisted surgery. Since additional hardware like tracking systems or the robot encoders are cumbersome and lack accuracy, surgical vision is evolving as promising techniques to segment and track the instruments using only the endoscopic images. However, what is missing so far are common image data sets for consistent evaluation and benchmarking of algorithms against each other. The paper presents a comparative validation study of different vision-based methods for instrument segmentation and tracking in the context of robotic as well as conventional laparoscopic surgery. The contribution of the paper is twofold: we introduce a comprehensive validation data set that was provided to the study participants and present the results of the comparative validation study. Based on the results of the validation study, we arrive at the conclusion that modern deep learning approaches outperform other methods in instrument segmentation tasks, but the results are still not perfect. Furthermore, we show that merging results from different methods actually significantly increases accuracy in comparison to the best stand-alone method. On the other hand, the results of the instrument tracking task show that this is still an open challenge, especially during challenging scenarios in conventional laparoscopic surgery.

Tobias Ross, David Zimmerer, Anant Vemuri et al.

Surgical data science is a new research field that aims to observe all aspects of the patient treatment process in order to provide the right assistance at the right time. Due to the breakthrough successes of deep learning-based solutions for automatic image annotation, the availability of reference annotations for algorithm training is becoming a major bottleneck in the field. The purpose of this paper was to investigate the concept of self-supervised learning to address this issue. Our approach is guided by the hypothesis that unlabeled video data can be used to learn a representation of the target domain that boosts the performance of state-of-the-art machine learning algorithms when used for pre-training. Core of the method is an auxiliary task based on raw endoscopic video data of the target domain that is used to initialize the convolutional neural network (CNN) for the target task. In this paper, we propose the re-colorization of medical images with a generative adversarial network (GAN)-based architecture as auxiliary task. A variant of the method involves a second pre-training step based on labeled data for the target task from a related domain. We validate both variants using medical instrument segmentation as target task. The proposed approach can be used to radically reduce the manual annotation effort involved in training CNNs. Compared to the baseline approach of generating annotated data from scratch, our method decreases exploratively the number of labeled images by up to 75% without sacrificing performance. Our method also outperforms alternative methods for CNN pre-training, such as pre-training on publicly available non-medical or medical data using the target task (in this instance: segmentation). As it makes efficient use of available (non-)public and (un-)labeled data, the approach has the potential to become a valuable tool for CNN (pre-)training.

Sebastian Bodenstedt, Martin Wagner, Darko Katić et al.

Computer-assisted surgery (CAS) aims to provide the surgeon with the right type of assistance at the right moment. Such assistance systems are especially relevant in laparoscopic surgery, where CAS can alleviate some of the drawbacks that surgeons incur. For many assistance functions, e.g. displaying the location of a tumor at the appropriate time or suggesting what instruments to prepare next, analyzing the surgical workflow is a prerequisite. Since laparoscopic interventions are performed via endoscope, the video signal is an obvious sensor modality to rely on for workflow analysis. Image-based workflow analysis tasks in laparoscopy, such as phase recognition, skill assessment, video indexing or automatic annotation, require a temporal distinction between video frames. Generally computer vision based methods that generalize from previously seen data are used. For training such methods, large amounts of annotated data are necessary. Annotating surgical data requires expert knowledge, therefore collecting a sufficient amount of data is difficult, time-consuming and not always feasible. In this paper, we address this problem by presenting an unsupervised method for training a convolutional neural network (CNN) to differentiate between laparoscopic video frames on a temporal basis. We extract video frames at regular intervals from 324 unlabeled laparoscopic interventions, resulting in a dataset of approximately 2.2 million images. From this dataset, we extract image pairs from the same video and train a CNN to determine their temporal order. To solve this problem, the CNN has to extract features that are relevant for comprehending laparoscopic workflow. Furthermore, we demonstrate that such a CNN can be adapted for surgical workflow segmentation. We performed image-based workflow segmentation on a publicly available dataset of 7 cholecystectomies and 9 colorectal interventions.