We propose a deep mixture of multimodal hierarchical variational auto-encoders called MMHVAE that synthesizes missing images from observed images in different modalities. MMHVAE's design focuses on tackling four challenges: (i) creating a complex latent representation of multimodal data to generate high-resolution images; (ii) encouraging the variational distributions to estimate the missing information needed for cross-modal image synthesis; (iii) learning to fuse multimodal information in the context of missing data; (iv) leveraging dataset-level information to handle incomplete data sets at training time. Extensive experiments are performed on the challenging problem of pre-operative brain multi-parametric magnetic resonance and intra-operative ultrasound imaging.

Mutual Information (MI) is a fundamental measure of statistical dependence widely used in representation learning. While direct optimization of MI via its definition as a Kullback-Leibler divergence (KLD) is often intractable, many recent methods have instead maximized alternative dependence measures, most notably, the Jensen-Shannon divergence (JSD) between joint and product of marginal distributions via discriminative losses. However, the connection between these surrogate objectives and MI remains poorly understood. In this work, we bridge this gap by deriving a new, tight, and tractable lower bound on KLD as a function of JSD in the general case. By specializing this bound to joint and marginal distributions, we demonstrate that maximizing the JSD-based information increases a guaranteed lower bound on mutual information. Furthermore, we revisit the practical implementation of JSD-based objectives and observe that minimizing the cross-entropy loss of a binary classifier trained to distinguish joint from marginal pairs recovers a known variational lower bound on the JSD. Extensive experiments demonstrate that our lower bound is tight when applied to MI estimation. We compared our lower bound to state-of-the-art neural estimators of variational lower bound across a range of established reference scenarios. Our lower bound estimator consistently provides a stable, low-variance estimate of a tight lower bound on MI. We also demonstrate its practical usefulness in the context of the Information Bottleneck framework. Taken together, our results provide new theoretical justifications and strong empirical evidence for using discriminative learning in MI-based representation learning.

We present in this paper a novel approach for 3D/2D intraoperative registration during neurosurgery via cross-modal inverse neural rendering. Our approach separates implicit neural representation into two components, handling anatomical structure preoperatively and appearance intraoperatively. This disentanglement is achieved by controlling a Neural Radiance Field's appearance with a multi-style hypernetwork. Once trained, the implicit neural representation serves as a differentiable rendering engine, which can be used to estimate the surgical camera pose by minimizing the dissimilarity between its rendered images and the target intraoperative image. We tested our method on retrospective patients' data from clinical cases, showing that our method outperforms state-of-the-art while meeting current clinical standards for registration. Code and additional resources can be found at https://maxfehrentz.github.io/style-ngp/.

We propose in this paper a texture-invariant 2D keypoints descriptor specifically designed for matching preoperative Magnetic Resonance (MR) images with intraoperative Ultrasound (US) images. We introduce a matching-by-synthesis strategy, where intraoperative US images are synthesized from MR images accounting for multiple MR modalities and intraoperative US variability. We build our training set by enforcing keypoints localization over all images then train a patient-specific descriptor network that learns texture-invariant discriminant features in a supervised contrastive manner, leading to robust keypoints descriptors. Our experiments on real cases with ground truth show the effectiveness of the proposed approach, outperforming the state-of-the-art methods and achieving 80.35% matching precision on average.

3D reconstruction of cerebral vasculature from 2D biplanar projections could significantly improve diagnosis and treatment planning. We introduce a novel approach to tackle this challenging task by initially backprojecting the two projections, a process that traditionally results in unsatisfactory outcomes due to inherent ambiguities. To overcome this, we employ a U-Net approach trained to resolve these ambiguities, leading to significant improvement in reconstruction quality. The process is further refined using a Maximum A Posteriori strategy with a prior that favors continuity, leading to enhanced 3D reconstructions. We evaluated our approach using a comprehensive dataset comprising segmentations from approximately 700 MR angiography scans, from which we generated paired realistic biplanar DRRs. Upon testing with held-out data, our method achieved an 80% Dice similarity w.r.t the ground truth, superior to existing methods. Our code and dataset are available at https://github.com/Wapity/3DBrainXVascular.

Automatic segmentation of vestibular schwannoma (VS) from routine clinical MRI has potential to improve clinical workflow, facilitate treatment decisions, and assist patient management. Previous work demonstrated reliable automatic segmentation performance on datasets of standardized MRI images acquired for stereotactic surgery planning. However, diagnostic clinical datasets are generally more diverse and pose a larger challenge to automatic segmentation algorithms, especially when post-operative images are included. In this work, we show for the first time that automatic segmentation of VS on routine MRI datasets is also possible with high accuracy. We acquired and publicly release a curated multi-center routine clinical (MC-RC) dataset of 160 patients with a single sporadic VS. For each patient up to three longitudinal MRI exams with contrast-enhanced T1-weighted (ceT1w) (n = 124) and T2-weighted (T2w) (n = 363) images were included and the VS manually annotated. Segmentations were produced and verified in an iterative process: (1) initial segmentations by a specialized company; (2) review by one of three trained radiologists; and (3) validation by an expert team. Inter- and intra-observer reliability experiments were performed on a subset of the dataset. A state-of-the-art deep learning framework was used to train segmentation models for VS. Model performance was evaluated on a MC-RC hold-out testing set, another public VS datasets, and a partially public dataset. The generalizability and robustness of the VS deep learning segmentation models increased significantly when trained on the MC-RC dataset. Dice similarity coefficients (DSC) achieved by our model are comparable to those achieved by trained radiologists in the inter-observer experiment. On the MC-RC testing set, median DSCs were 86.2(9.5) for ceT1w, 89.4(7.0) for T2w, and 86.4(8.6) for combined ceT1w+T2w input images. On another public dataset acquired for Gamma Knife stereotactic radiosurgery our model achieved median DSCs of 95.3(2.9), 92.8(3.8), and 95.5(3.3), respectively. In contrast, models trained on the Gamma Knife dataset did not generalize well as illustrated by significant underperformance on the MC-RC routine MRI dataset, highlighting the importance of data variability in the development of robust VS segmentation models. The MC-RC dataset and all trained deep learning models were made available online.

Intraoperative ultrasound (iUS) imaging has the potential to improve surgical outcomes in brain surgery. However, its interpretation is challenging, even for expert neurosurgeons. In this work, we designed the first patient-specific framework that performs brain tumor segmentation in trackerless iUS. To disambiguate ultrasound imaging and adapt to the neurosurgeon's surgical objective, a patient-specific real-time network is trained using synthetic ultrasound data generated by simulating virtual iUS sweep acquisitions in pre-operative MR data. Extensive experiments performed in real ultrasound data demonstrate the effectiveness of the proposed approach, allowing for adapting to the surgeon's definition of surgical targets and outperforming non-patient-specific models, neurosurgeon experts, and high-end tracking systems. Our code is available at: https://github.com/ReubenDo/MHVAE- Seg.

OBJECTIVES:The shape is commonly used to describe the objects. State-of-the-art algorithms in medical imaging are predominantly diverging from computer vision, where voxel grids, meshes, point clouds, and implicit surface models are used. This is seen from the growing popularity of ShapeNet (51,300 models) and Princeton ModelNet (127,915 models). However, a large collection of anatomical shapes (e.g., bones, organs, vessels) and 3D models of surgical instruments is missing. METHODS:We present MedShapeNet to translate data-driven vision algorithms to medical applications and to adapt state-of-the-art vision algorithms to medical problems. As a unique feature, we directly model the majority of shapes on the imaging data of real patients. We present use cases in classifying brain tumors, skull reconstructions, multi-class anatomy completion, education, and 3D printing. RESULTS:By now, MedShapeNet includes 23 datasets with more than 100,000 shapes that are paired with annotations (ground truth). Our data is freely accessible via a web interface and a Python application programming interface and can be used for discriminative, reconstructive, and variational benchmarks as well as various applications in virtual, augmented, or mixed reality, and 3D printing. CONCLUSIONS:MedShapeNet contains medical shapes from anatomy and surgical instruments and will continue to collect data for benchmarks and applications. The project page is: https://medshapenet.ikim.nrw/.

Accurate assessment of lymph node size in 3D CT scans is crucial for cancer staging, therapeutic management, and monitoring treatment response. Existing state-of-the-art segmentation frameworks in medical imaging often rely on fully annotated datasets. However, for lymph node segmentation, these datasets are typically small due to the extensive time and expertise required to annotate the numerous lymph nodes in 3D CT scans. Weakly-supervised learning, which leverages incomplete or noisy annotations, has recently gained interest in the medical imaging community as a potential solution. Despite the variety of weakly-supervised techniques proposed, most have been validated only on private datasets or small publicly available datasets. To address this limitation, the Mediastinal Lymph Node Quantification (LNQ) challenge was organized in conjunction with the 26th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2023). This challenge aimed to advance weakly-supervised segmentation methods by providing a new, partially annotated dataset and a robust evaluation framework. A total of 16 teams from 5 countries submitted predictions to the validation leaderboard, and 6 teams from 3 countries participated in the evaluation phase. The results highlighted both the potential and the current limitations of weakly-supervised approaches. On one hand, weakly-supervised approaches obtained relatively good performance with a median Dice score of 61.0%. On the other hand, top-ranked teams, with a median Dice score exceeding 70%, boosted their performance by leveraging smaller but fully annotated datasets to combine weak supervision and full supervision. This highlights both the promise of weakly-supervised methods and the ongoing need for high-quality, fully annotated data to achieve higher segmentation performance.

Whole brain parcellation requires inferring hundreds of segmentation labels in large image volumes and thus presents significant practical challenges for deep learning approaches. We introduce label merge-and-split, a method that first greatly reduces the effective number of labels required for learning-based whole brain parcellation and then recovers original labels. Using a greedy graph colouring algorithm, our method automatically groups and merges multiple spatially separate labels prior to model training and inference. The merged labels may be semantically unrelated. A deep learning model is trained to predict merged labels. At inference time, original labels are restored using atlas-based influence regions. In our experiments, the proposed approach reduces the number of labels by up to 68% while achieving segmentation accuracy comparable to the baseline method without label merging and splitting. Moreover, model training and inference times as well as GPU memory requirements were reduced significantly. The proposed method can be applied to all semantic segmentation tasks with a large number of spatially separate classes within an atlas-based prior.

The standard of care for brain tumors is maximal safe surgical resection. Neuronavigation augments the surgeon's ability to achieve this but loses validity as surgery progresses due to brain shift. Moreover, gliomas are often indistinguishable from surrounding healthy brain tissue. Intraoperative magnetic resonance imaging (iMRI) and ultrasound (iUS) help visualize the tumor and brain shift. iUS is faster and easier to incorporate into surgical workflows but offers a lower contrast between tumorous and healthy tissues than iMRI. With the success of data-hungry Artificial Intelligence algorithms in medical image analysis, the benefits of sharing well-curated data cannot be overstated. To this end, we provide the largest publicly available MRI and iUS database of surgically treated brain tumors, including gliomas (n=92), metastases (n=11), and others (n=11). This collection contains 369 preoperative MRI series, 320 3D iUS series, 301 iMRI series, and 356 segmentations collected from 114 consecutive patients at a single institution. This database is expected to help brain shift and image analysis research and neurosurgical training in interpreting iUS and iMRI.

INTRODUCTION: Automated delineation of gliomas in intraoperative ultrasound (iUS) could provide timely information to surgeons. While artificial intelligence (AI) frameworks have shown promising results for delineating gliomas in preoperative MRI, automated delineation in iUS remains a challenging and open problem. METHODS: A deep learning algorithm was developed to automatically delineate gliomas in intraoperative ultrasound. Instead of using time-consuming manual delineations from experts, the algorithm was developed using MR-based automated delineations. First, a MR nnUnet was trained for delineating gliomas in 3D preoperative MRI using an annotated public dataset (N=1000; RSNA-ASNR-MICCAI-BraTS 2021). Then, this algorithm delineated gliomas in preoperative MRI of patients (N = 84) in whom pre-dural opening 3D iUS reconstructed from a tracked handheld 2D probe was also acquired. To obtain the MR-based delineations in iUS, preoperative MRI and iUS were aligned using NiftyReg. Finally, a 3D iUS nnUnet was trained using a random subset (N = 78) of this pseudo-annotated iUS dataset. The method was assessed on the remaining test set (N = 943 slices from 6 volumes), with a treating neurosurgeon delineating gliomas in 2D iUS only and then in 2D iUS and aligned MR slices to obtain gold-standard delineations. RESULTS: The median Dice scores of the nnUnet and the neurosurgeon were respectively 78.6% (IQ:15.5%) and 84.3% (IQ: 21.1%) against the gold-standard delineations. In this preliminary study, the network's performance was comparable to the neurosurgeon's (two paired one-sided tests, equivalence margin:5%, p<1e-5, N = 943), with rapid delineation (<10 second) of 3D iUS (95-216 slices) compared to neurogeons (32-93 minutes), making automated delineation practical during surgery. CONCLUSIONS: The results demonstrate the feasibility of using AI-based methods for automated delineation of brain tumors in iUS, which could improve surgical outcomes by providing real-time information to surgeons.

Although Digital Subtraction Angiography (DSA) is the most important imaging for visualizing cerebrovascular anatomy, its interpretation by clinicians remains difficult. This is particularly true when treating arteriovenous malformations (AVMs), where entangled vasculature connecting arteries and veins needs to be carefully identified.The presented method aims to enhance DSA image series by highlighting critical information via automatic classification of vessels using a combination of two learning models: An unsupervised machine learning method based on Independent Component Analysis that decomposes the phases of flow and a convolutional neural network that automatically delineates the vessels in image space. The proposed method was tested on clinical DSA images series and demonstrated efficient differentiation between arteries and veins that provides a viable solution to enhance visualizations for clinical use.

Surgical instrument segmentation is recognised as a key enabler to provide advanced surgical assistance and improve computer assisted interventions. In this work, we propose SegMatch, a semi supervised learning method to reduce the need for expensive annotation for laparoscopic and robotic surgical images. SegMatch builds on FixMatch, a widespread semi supervised classification pipeline combining consistency regularization and pseudo labelling, and adapts it for the purpose of segmentation. In our proposed SegMatch, the unlabelled images are weakly augmented and fed into the segmentation model to generate a pseudo-label to enforce the unsupervised loss against the output of the model for the adversarial augmented image on the pixels with a high confidence score. Our adaptation for segmentation tasks includes carefully considering the equivariance and invariance properties of the augmentation functions we rely on. To increase the relevance of our augmentations, we depart from using only handcrafted augmentations and introduce a trainable adversarial augmentation strategy. Our algorithm was evaluated on the MICCAI Instrument Segmentation Challenge datasets Robust-MIS 2019 and EndoVis 2017. Our results demonstrate that adding unlabelled data for training purposes allows us to surpass the performance of fully supervised approaches which are limited by the availability of training data in these challenges. SegMatch also outperforms a range of state-of-the-art semi-supervised learning semantic segmentation models in different labelled to unlabelled data ratios.

Recent advances in medical Deep Learning (DL) have enabled the significant reduction in time required to extract anatomical segmentations from 3-Dimensional images in an unprecedented manner. Among these methods, supervised segmentation-based approaches using variations of the UNet architecture remain extremely popular. However, these methods remain tied to the input images' resolution, and their generalisation performance relies heavily on the data distribution over the training dataset. Recently, a new family of approaches based on 3D geometric DL has emerged. These approaches encompass both implicit and explicit surface representation methods and promises to represent a 3D volume using a continuous representation of its surface whilst conserving its topological properties. It has been conjectured that these geometrical methods are more robust to out-of-distribution data and have increased generalisation properties. In this paper, we test these hypotheses for the challenging task of cortical surface reconstruction (CSR) using recently proposed architectures.

We introduce MHVAE, a deep hierarchical variational auto-encoder (VAE) that synthesizes missing images from various modalities. Extending multi-modal VAEs with a hierarchical latent structure, we introduce a probabilistic formulation for fusing multi-modal images in a common latent representation while having the flexibility to handle incomplete image sets as input. Moreover, adversarial learning is employed to generate sharper images. Extensive experiments are performed on the challenging problem of joint intra-operative ultrasound (iUS) and Magnetic Resonance (MR) synthesis. Our model outperformed multi-modal VAEs, conditional GANs, and the current state-of-the-art unified method (ResViT) for synthesizing missing images, demonstrating the advantage of using a hierarchical latent representation and a principled probabilistic fusion operation. Our code is publicly available ( https://github.com/ReubenDo/MHVAE ).

We present a novel method for intraoperative patient-to-image registration by learning Expected Appearances. Our method uses preoperative imaging to synthesize patient-specific expected views through a surgical microscope for a predicted range of transformations. Our method estimates the camera pose by minimizing the dissimilarity between the intraoperative 2D view through the optical microscope and the synthesized expected texture. In contrast to conventional methods, our approach transfers the processing tasks to the preoperative stage, reducing thereby the impact of low-resolution, distorted, and noisy intraoperative images, that often degrade the registration accuracy. We applied our method in the context of neuronavigation during brain surgery. We evaluated our approach on synthetic data and on retrospective data from 6 clinical cases. Our method outperformed state-of-the-art methods and achieved accuracies that met current clinical standards.

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.

Objectives Surgical planning of vestibular schwannoma surgery would benefit greatly from a robust method of delineating the facial-vestibulocochlear nerve complex with respect to the tumour. This study aimed to optimise a multi-shell readout-segmented diffusion-weighted imaging (rs-DWI) protocol and develop a novel post-processing pipeline to delineate the facial-vestibulocochlear complex within the skull base region, evaluating its accuracy intraoperatively using neuronavigation and tracked electrophysiological recordings. Methods In a prospective study of five healthy volunteers and five patients who underwent vestibular schwannoma surgery, rs-DWI was performed and colour tissue maps (CTM) and probabilistic tractography of the cranial nerves were generated. In patients, the average symmetric surface distance (ASSD) and 95% Hausdorff distance (HD-95) were calculated with reference to the neuroradiologist-approved facial nerve segmentation. The accuracy of patient results was assessed intraoperatively using neuronavigation and tracked electrophysiological recordings. Results Using CTM alone, the facial-vestibulocochlear complex of healthy volunteer subjects was visualised on 9/10 sides. CTM were generated in all 5 patients with vestibular schwannoma enabling the facial nerve to be accurately identified preoperatively. The mean ASSD between the annotators’ two segmentations was 1.11 mm (SD 0.40) and the mean HD-95 was 4.62 mm (SD 1.78). The median distance from the nerve segmentation to a positive stimulation point was 1.21 mm (IQR 0.81–3.27 mm) and 2.03 mm (IQR 0.99–3.84 mm) for the two annotators, respectively. Conclusions rs-DWI may be used to acquire dMRI data of the cranial nerves within the posterior fossa. Clinical relevance statement Readout-segmented diffusion-weighted imaging and colour tissue mapping provide 1–2 mm spatially accurate imaging of the facial-vestibulocochlear nerve complex, enabling accurate preoperative localisation of the facial nerve. This study evaluated the technique in 5 healthy volunteers and 5 patients with vestibular schwannoma. Key Points • Readout-segmented diffusion-weighted imaging (rs-DWI) with colour tissue mapping (CTM) visualised the facial-vestibulocochlear nerve complex on 9/10 sides in 5 healthy volunteer subjects. • Using rs-DWI and CTM, the facial nerve was visualised in all 5 patients with vestibular schwannoma and within 1.21–2.03 mm of the nerve’s true intraoperative location. • Reproducible results were obtained on different scanners.

Registering CT images of the chest is a crucial step for several tasks such as disease progression tracking or surgical planning. It is also a challenging step because of the heterogeneous content of the human abdomen which implies complex deformations. In this work, we focus on accurately registering a subset of organs of interest. We register organ surface point clouds, as may typically be extracted from an automatic segmentation pipeline, by expanding the Bayesian Coherent Point Drift algorithm (BCPD). We introduce MO-BCPD, a multi-organ version of the BCPD algorithm which explicitly models three important aspects of this task: organ individual elastic properties, inter-organ motion coherence and segmentation inaccuracy. This model also provides an interpolation framework to estimate the deformation of the entire volume. We demonstrate the efficiency of our method by registering different patients from the LITS challenge dataset. The target registration error on anatomical landmarks is almost twice as small for MO-BCPD compared to standard BCPD while imposing the same constraints on individual organs deformation.