Polina Golland

Nalini M. Singh, Neel Dey, Malte Hoffmann et al. · mit

Motion artifacts are a pervasive problem in MRI, leading to misdiagnosis or mischaracterization in population-level imaging studies. Current retrospective rigid intra-slice motion correction techniques jointly optimize estimates of the image and the motion parameters. In this paper, we use a deep network to reduce the joint image-motion parameter search to a search over rigid motion parameters alone. Our network produces a reconstruction as a function of two inputs: corrupted k-space data and motion parameters. We train the network using simulated, motion-corrupted k-space data generated with known motion parameters. At test-time, we estimate unknown motion parameters by minimizing a data consistency loss between the motion parameters, the network-based image reconstruction given those parameters, and the acquired measurements. Intra-slice motion correction experiments on simulated and realistic 2D fast spin echo brain MRI achieve high reconstruction fidelity while providing the benefits of explicit data consistency optimization. Our code is publicly available at https://www.github.com/nalinimsingh/neuroMoCo.

Vivek Gopalakrishnan, Polina Golland · mit

The use of digitally reconstructed radiographs (DRRs) to solve inverse problems such as slice-to-volume registration and 3D reconstruction is well-studied in preoperative settings. In intraoperative imaging, the utility of DRRs is limited by the challenges in generating them in real-time and supporting optimization procedures that rely on repeated DRR synthesis. While immense progress has been made in accelerating the generation of DRRs through algorithmic refinements and GPU implementations, DRR-based optimization remains slow because most DRR generators do not offer a straightforward way to obtain gradients with respect to the imaging parameters. To make DRRs interoperable with gradient-based optimization and deep learning frameworks, we have reformulated Siddon's method, the most popular ray-tracing algorithm used in DRR generation, as a series of vectorized tensor operations. We implemented this vectorized version of Siddon's method in PyTorch, taking advantage of the library's strong automatic differentiation engine to make this DRR generator fully differentiable with respect to its parameters. Additionally, using GPU-accelerated tensor computation enables our vectorized implementation to achieve rendering speeds equivalent to state-of-the-art DRR generators implemented in CUDA and C++. We illustrate the resulting method in the context of slice-to-volume registration. Moreover, our simulations suggest that the loss landscapes for the slice-to-volume registration problem are convex in the neighborhood of the optimal solution, and gradient-based registration promises a much faster solution than prevailing gradient-free optimization strategies. The proposed DRR generator enables fast computer vision algorithms to support image guidance in minimally invasive procedures. Our implementation is publically available at https://github.com/v715/DiffDRR.

Neel Dey, S. Mazdak Abulnaga, Benjamin Billot et al. · mit

Star-convex shapes arise across bio-microscopy and radiology in the form of nuclei, nodules, metastases, and other units. Existing instance segmentation networks for such structures train on densely labeled instances for each dataset, which requires substantial and often impractical manual annotation effort. Further, significant reengineering or finetuning is needed when presented with new datasets and imaging modalities due to changes in contrast, shape, orientation, resolution, and density. We present AnyStar, a domain-randomized generative model that simulates synthetic training data of blob-like objects with randomized appearance, environments, and imaging physics to train general-purpose star-convex instance segmentation networks. As a result, networks trained using our generative model do not require annotated images from unseen datasets. A single network trained on our synthesized data accurately 3D segments C. elegans and P. dumerilii nuclei in fluorescence microscopy, mouse cortical nuclei in micro-CT, zebrafish brain nuclei in EM, and placental cotyledons in human fetal MRI, all without any retraining, finetuning, transfer learning, or domain adaptation. Code is available at https://github.com/neel-dey/AnyStar.

S. Mazdak Abulnaga, Sean I. Young, Katherine Hobgood et al. · mit

Blood oxygen level dependent (BOLD) MRI with maternal hyperoxia can assess oxygen transport within the placenta and has emerged as a promising tool to study placental function. Measuring signal changes over time requires segmenting the placenta in each volume of the time series. Due to the large number of volumes in the BOLD time series, existing studies rely on registration to map all volumes to a manually segmented template. As the placenta can undergo large deformation due to fetal motion, maternal motion, and contractions, this approach often results in a large number of discarded volumes, where the registration approach fails. In this work, we propose a machine learning model based on a U-Net neural network architecture to automatically segment the placenta in BOLD MRI and apply it to segmenting each volume in a time series. We use a boundary-weighted loss function to accurately capture the placental shape. Our model is trained and tested on a cohort of 91 subjects containing healthy fetuses, fetuses with fetal growth restriction, and mothers with high BMI. We achieve a Dice score of 0.83+/-0.04 when matching with ground truth labels and our model performs reliably in segmenting volumes in both normoxic and hyperoxic points in the BOLD time series. Our code and trained model are available at https://github.com/mabulnaga/automatic-placenta-segmentation.

Zeen Chi, Zhongxiao Cong, Clinton J. Wang et al. · mit

We present a method for fast biomedical image atlas construction using neural fields. Atlases are key to biomedical image analysis tasks, yet conventional and deep network estimation methods remain time-intensive. In this preliminary work, we frame subject-specific atlas building as learning a neural field of deformable spatiotemporal observations. We apply our method to learning subject-specific atlases and motion stabilization of dynamic BOLD MRI time-series of fetuses in utero. Our method yields high-quality atlases of fetal BOLD time-series with $\sim$5-7$\times$ faster convergence compared to existing work. While our method slightly underperforms well-tuned baselines in terms of anatomical overlap, it estimates templates significantly faster, thus enabling rapid processing and stabilization of large databases of 4D dynamic MRI acquisitions. Code is available at https://github.com/Kidrauh/neural-atlasing

Sebo Diaz, Polina Golland, Elfar Adalsteinsson et al. · mit

We present DropGen, a simple and theoretically-grounded approach for domain generalization in 3D biomedical image segmentation. Modern segmentation models degrade sharply under shifts in modality, disease severity, clinical sites, and other factors, creating brittle models that limit reliable deployment. Existing domain generalization methods rely on extreme augmentations, mixing domain statistics, or architectural redesigns, yet incur significant implementation overhead and yield inconsistent performance across biomedical settings. DropGen instead proposes a principled learning strategy with minimal overhead that leverages both source-domain image intensities and domain-stable foundation model representations to train robust segmentation models. As a result, DropGen achieves strong gains in both fully supervised and few-shot segmentation across a broad range of shifts in biomedical studies. Unlike prior approaches, DropGen is architecture- and loss-agnostic, compatible with standard augmentation pipelines, computationally lightweight, and tackles arbitrary anatomical regions. Our implementation is freely available at https://github.com/sebodiaz/DropGen.

Yingcheng Liu, Neerav Karani, Neel Dey et al. · mit

The placenta plays a crucial role in fetal development. Automated 3D placenta segmentation from fetal EPI MRI holds promise for advancing prenatal care. This paper proposes an effective semi-supervised learning method for improving placenta segmentation in fetal EPI MRI time series. We employ consistency regularization loss that promotes consistency under spatial transformation of the same image and temporal consistency across nearby images in a time series. The experimental results show that the method improves the overall segmentation accuracy and provides better performance for outliers and hard samples. The evaluation also indicates that our method improves the temporal coherency of the prediction, which could lead to more accurate computation of temporal placental biomarkers. This work contributes to the study of the placenta and prenatal clinical decision-making. Code is available at https://github.com/firstmover/cr-seg.

Clinton J. Wang, Polina Golland · mit

With the emergence of powerful representations of continuous data in the form of neural fields, there is a need for discretization invariant learning: an approach for learning maps between functions on continuous domains without being sensitive to how the function is sampled. We present a new framework for understanding and designing discretization invariant neural networks (DI-Nets), which generalizes many discrete networks such as convolutional neural networks as well as continuous networks such as neural operators. Our analysis establishes upper bounds on the deviation in model outputs under different finite discretizations, and highlights the central role of point set discrepancy in characterizing such bounds. This insight leads to the design of a family of neural networks driven by numerical integration via quasi-Monte Carlo sampling with discretizations of low discrepancy. We prove by construction that DI-Nets universally approximate a large class of maps between integrable function spaces, and show that discretization invariance also describes backpropagation through such models. Applied to neural fields, convolutional DI-Nets can learn to classify and segment visual data under various discretizations, and sometimes generalize to new types of discretizations at test time. Code: https://github.com/clintonjwang/DI-net.

Junshen Xu, Daniel Moyer, P. Ellen Grant et al. · mit

Volumetric reconstruction of fetal brains from multiple stacks of MR slices, acquired in the presence of almost unpredictable and often severe subject motion, is a challenging task that is highly sensitive to the initialization of slice-to-volume transformations. We propose a novel slice-to-volume registration method using Transformers trained on synthetically transformed data, which model multiple stacks of MR slices as a sequence. With the attention mechanism, our model automatically detects the relevance between slices and predicts the transformation of one slice using information from other slices. We also estimate the underlying 3D volume to assist slice-to-volume registration and update the volume and transformations alternately to improve accuracy. Results on synthetic data show that our method achieves lower registration error and better reconstruction quality compared with existing state-of-the-art methods. Experiments with real-world MRI data are also performed to demonstrate the ability of the proposed model to improve the quality of 3D reconstruction under severe fetal motion.

Clinton J. Wang, Polina Golland · mit

One little-explored frontier of image generation and editing is the task of interpolating between two input images, a feature missing from all currently deployed image generation pipelines. We argue that such a feature can expand the creative applications of such models, and propose a method for zero-shot interpolation using latent diffusion models. We apply interpolation in the latent space at a sequence of decreasing noise levels, then perform denoising conditioned on interpolated text embeddings derived from textual inversion and (optionally) subject poses. For greater consistency, or to specify additional criteria, we can generate several candidates and use CLIP to select the highest quality image. We obtain convincing interpolations across diverse subject poses, image styles, and image content, and show that standard quantitative metrics such as FID are insufficient to measure the quality of an interpolation. Code and data are available at https://clintonjwang.github.io/interpolation.

Neerav Karani, Neel Dey, Polina Golland · mit

Neural network prediction probabilities and accuracy are often only weakly-correlated. Inherent label ambiguity in training data for image segmentation aggravates such miscalibration. We show that logit consistency across stochastic transformations acts as a spatially varying regularizer that prevents overconfident predictions at pixels with ambiguous labels. Our boundary-weighted extension of this regularizer provides state-of-the-art calibration for prostate and heart MRI segmentation.

Ramya Muthukrishnan, Borjan Gagoski, Aryn Lee et al.

We present E(3)-Pose, a novel fast pose estimation method that jointly and explicitly models rotation equivariance and object symmetry. Our work is motivated by the challenging problem of accounting for fetal head motion during a diagnostic MRI scan. We aim to enable automatic adaptive prescription of 2D diagnostic MRI slices with 6-DoF head pose estimation, supported by 3D MRI volumes rapidly acquired before each 2D slice. Existing methods struggle to generalize to clinical volumes, due to pose ambiguities induced by inherent anatomical symmetries, as well as low resolution, noise, and artifacts. In contrast, E(3)-Pose captures anatomical symmetries and rigid pose equivariance by construction, and yields robust estimates of the fetal head pose. Our experiments on publicly available and representative clinical fetal MRI datasets demonstrate the superior robustness and generalization of our method across domains. Crucially, E(3)-Pose achieves state-of-the-art accuracy on clinical MRI volumes, supporting future clinical translation. Our implementation is publicly available at github.com/MedicalVisionGroup/E3-Pose.

Dominik Hollidt, Clinton Wang, Polina Golland et al. · mit

We present a novel approach to perform 3D semantic segmentation solely from 2D supervision by leveraging Neural Radiance Fields (NeRFs). By extracting features along a surface point cloud, we achieve a compact representation of the scene which is sample-efficient and conducive to 3D reasoning. Learning this feature space in an unsupervised manner via masked autoencoding enables few-shot segmentation. Our method is agnostic to the scene parameterization, working on scenes fit with any type of NeRF.

Yiming Chen, Niharika S. D'Souza, Akshith Mandepally et al. · mit

Precisely estimating lumen boundaries in intravascular ultrasound (IVUS) is needed for sizing interventional stents to treat deep vein thrombosis (DVT). Unfortunately, current segmentation networks like the UNet lack the precision needed for clinical adoption in IVUS workflows. This arises due to the difficulty of automatically learning accurate lumen contour from limited training data while accounting for the radial geometry of IVUS imaging. We propose the Geo-UNet framework to address these issues via a design informed by the geometry of the lumen contour segmentation task. We first convert the input data and segmentation targets from Cartesian to polar coordinates. Starting from a convUNet feature extractor, we propose a two-task setup, one for conventional pixel-wise labeling and the other for single boundary lumen-contour localization. We directly combine the two predictions by passing the predicted lumen contour through a new activation (named CDFeLU) to filter out spurious pixel-wise predictions. Our unified loss function carefully balances area-based, distance-based, and contour-based penalties to provide near clinical-grade generalization in unseen patient data. We also introduce a lightweight, inference-time technique to enhance segmentation smoothness. The efficacy of our framework on a venous IVUS dataset is shown against state-of-the-art models.

Deborah Pereg, Martin Villiger, Brett Bouma et al.

The statistical supervised learning framework assumes an input-output set with a joint probability distribution that is reliably represented by the training dataset. The learner is then required to output a prediction rule learned from the training dataset's input-output pairs. In this work, we provide meaningful insights into the asymptotic equipartition property (AEP) \citep{Shannon:1948} in the context of machine learning, and illuminate some of its potential ramifications for few-shot learning. We provide theoretical guarantees for reliable learning under the information-theoretic AEP, and for the generalization error with respect to the sample size. We then focus on a highly efficient recurrent neural net (RNN) framework and propose a reduced-entropy algorithm for few-shot learning. We also propose a mathematical intuition for the RNN as an approximation of a sparse coding solver. We verify the applicability, robustness, and computational efficiency of the proposed approach with image deblurring and optical coherence tomography (OCT) speckle suppression. Our experimental results demonstrate significant potential for improving learning models' sample efficiency, generalization, and time complexity, that can therefore be leveraged for practical real-time applications.

Peiqi Wang, William M. Wells, Seth Berkowitz et al.

Image-text multimodal representation learning aligns data across modalities and enables important medical applications, e.g., image classification, visual grounding, and cross-modal retrieval. In this work, we establish a connection between multimodal representation learning and multiple instance learning. Based on this connection, we propose a generic framework for constructing permutation-invariant score functions with many existing multimodal representation learning approaches as special cases. Furthermore, we use the framework to derive a novel contrastive learning approach and demonstrate that our method achieves state-of-the-art results in several downstream tasks.

Keegan Quigley, Miriam Cha, Ruizhi Liao et al.

Automated analysis of chest radiography using deep learning has tremendous potential to enhance the clinical diagnosis of diseases in patients. However, deep learning models typically require large amounts of annotated data to achieve high performance -- often an obstacle to medical domain adaptation. In this paper, we build a data-efficient learning framework that utilizes radiology reports to improve medical image classification performance with limited labeled data (fewer than 1000 examples). Specifically, we examine image-captioning pretraining to learn high-quality medical image representations that train on fewer examples. Following joint pretraining of a convolutional encoder and transformer decoder, we transfer the learned encoder to various classification tasks. Averaged over 9 pathologies, we find that our model achieves higher classification performance than ImageNet-supervised and in-domain supervised pretraining when labeled training data is limited.

Peiqi Wang, Yingcheng Liu, Ching-Yun Ko et al.

Self-supervised representation learning on image-text data facilitates crucial medical applications, such as image classification, visual grounding, and cross-modal retrieval. One common approach involves contrasting semantically similar (positive) and dissimilar (negative) pairs of data points. Drawing negative samples uniformly from the training data set introduces false negatives, i.e., samples that are treated as dissimilar but belong to the same class. In healthcare data, the underlying class distribution is nonuniform, implying that false negatives occur at a highly variable rate. To improve the quality of learned representations, we develop a novel approach that corrects for false negatives. Our method can be viewed as a variant of debiased contrastive learning that uses estimated sample-specific class probabilities. We provide theoretical analysis of the objective function and demonstrate the proposed approach on both image and paired image-text data sets. Our experiments illustrate empirical advantages of sample-specific debiasing.

Margherita Firenze, Sean I. Young, Clinton J. Wang et al.

Fully convolutional networks have become the backbone of modern medical imaging due to their ability to learn multi-scale representations and perform end-to-end inference. Yet their potential for slice-to-volume reconstruction (SVR), the task of jointly estimating 3D anatomy and slice poses from misaligned 2D acquisitions, remains underexplored. We introduce a fast convolutional framework that fuses multiple orthogonal 2D slice stacks to recover coherent 3D structure and refines slice alignment through lightweight model-based optimization. Applied to fetal brain MRI, our approach reconstructs high-quality 3D volumes in under 10s, with 1s slice registration and accuracy on par with state-of-the-art iterative SVR pipelines, offering more than speedup. The framework uses non-rigid displacement fields to represent transformations, generalizing to other SVR problems like fetal body and placental MRI. Additionally, the fast inference time paves the way for real-time, scanner-side volumetric feedback during MRI acquisition.

Paul Weiser, Georg Langs, Wolfgang Bogner et al.

Introduction: Altered neurometabolism is an important pathological mechanism in many neurological diseases and brain cancer, which can be mapped non-invasively by Magnetic Resonance Spectroscopic Imaging (MRSI). Advanced MRSI using non-cartesian compressed-sense acquisition enables fast high-resolution metabolic imaging but has lengthy reconstruction times that limits throughput and needs expert user interaction. Here, we present a robust and efficient Deep Learning reconstruction to obtain high-quality metabolic maps. Methods: Fast high-resolution whole-brain metabolic imaging was performed at 3.4 mm$^3$ isotropic resolution with acquisition times between 4:11-9:21 min:s using ECCENTRIC pulse sequence on a 7T MRI scanner. Data were acquired in a high-resolution phantom and 27 human participants, including 22 healthy volunteers and 5 glioma patients. A deep neural network using recurring interlaced convolutional layers with joint dual-space feature representation was developed for deep learning ECCENTRIC reconstruction (Deep-ER). 21 subjects were used for training and 6 subjects for testing. Deep-ER performance was compared to conventional iterative Total Generalized Variation reconstruction using image and spectral quality metrics. Results: Deep-ER demonstrated 600-fold faster reconstruction than conventional methods, providing improved spatial-spectral quality and metabolite quantification with 12%-45% (P<0.05) higher signal-to-noise and 8%-50% (P<0.05) smaller Cramer-Rao lower bounds. Metabolic images clearly visualize glioma tumor heterogeneity and boundary. Conclusion: Deep-ER provides efficient and robust reconstruction for sparse-sampled MRSI. The accelerated acquisition-reconstruction MRSI is compatible with high-throughput imaging workflow. It is expected that such improved performance will facilitate basic and clinical MRSI applications.

Jian Wang, Razieh Faghihpirayesh, Danny Joca et al.

Deep learning has shown significant value in image registration, however, current techniques are either limited by the type and range of motion they can handle, or require iterative inference and/or retraining for new imaging data. To address these limitations, we introduce UniMo, a Unified Motion Correction framework that leverages deep neural networks to correct diverse motion in medical imaging. UniMo employs an alternating optimization scheme for a unified loss function to train an integrated model of 1) an equivariant neural network for global rigid motion correction and 2) an encoder-decoder network for local deformations. It features a geometric deformation augmenter that 1) enhances the robustness of global correction by addressing local deformations from non-rigid motion or geometric distortions, and 2) generates augmented data to improve training. UniMo is a hybrid model that uses both image intensities and shapes to achieve robust performance amid appearance variations, and therefore generalizes to multiple imaging modalities without retraining. We trained and tested UniMo to track motion in fetal magnetic resonance imaging, a challenging application due to 1) both large rigid and non-rigid motion, and 2) wide variations in image appearance. We then evaluated the trained model, without retraining, on MedMNIST, lung CT, and BraTS datasets. Results show that UniMo surpassed existing motion correction methods in accuracy, and notably enabled one-time training on a single modality while maintaining high stability and adaptability across unseen datasets. By offering a unified solution to motion correction, UniMo marks a significant advance in medical imaging, especially in applications with combined bulk and local motion. The code is available at: https://github.com/IntelligentImaging/UNIMO

Vivek Gopalakrishnan, Neel Dey, Polina Golland · mit

Surgical decisions are informed by aligning rapid portable 2D intraoperative images (e.g., X-rays) to a high-fidelity 3D preoperative reference scan (e.g., CT). 2D/3D image registration often fails in practice: conventional optimization methods are prohibitively slow and susceptible to local minima, while neural networks trained on small datasets fail on new patients or require impractical landmark supervision. We present DiffPose, a self-supervised approach that leverages patient-specific simulation and differentiable physics-based rendering to achieve accurate 2D/3D registration without relying on manually labeled data. Preoperatively, a CNN is trained to regress the pose of a randomly oriented synthetic X-ray rendered from the preoperative CT. The CNN then initializes rapid intraoperative test-time optimization that uses the differentiable X-ray renderer to refine the solution. Our work further proposes several geometrically principled methods for sampling camera poses from $\mathbf{SE}(3)$, for sparse differentiable rendering, and for driving registration in the tangent space $\mathfrak{se}(3)$ with geodesic and multiscale locality-sensitive losses. DiffPose achieves sub-millimeter accuracy across surgical datasets at intraoperative speeds, improving upon existing unsupervised methods by an order of magnitude and even outperforming supervised baselines. Our code is available at https://github.com/eigenvivek/DiffPose.

Keegan Quigley, Miriam Cha, Josh Barua et al.

Vision-language pretraining has been shown to produce high-quality visual encoders which transfer efficiently to downstream computer vision tasks. Contrastive learning approaches have increasingly been adopted for medical vision language pretraining (MVLP), yet recent developments in generative AI offer new modeling alternatives. This paper introduces RadTex, a CNN-encoder transformer-decoder architecture optimized for radiology. We explore bidirectional captioning as an alternative MVLP strategy and demonstrate that RadTex's captioning pretraining is competitive with established contrastive methods, achieving a CheXpert macro-AUC of 89.4%. Additionally, RadTex's lightweight text decoder not only generates clinically relevant radiology reports (macro-F1 score of 0.349), but also provides targeted, interactive responses, highlighting the utility of bidirectional captioning in advancing medical image analysis.

Athena Taymourtash, S. Mazdak Abulnaga, Esra Abaci Turk et al.

Establishing dense volumetric correspondences across anatomical shapes is essential for group-level analysis but remains challenging for implicit neural representations. Most existing implicit registration methods rely on supervision near the zero-level set and thus capture only surface correspondences, leaving interior deformations under-constrained. We introduce a volumetrically consistent implicit model that couples reconstruction of signed distance functions (SDFs) with neural diffeomorphic flow to learn a shared canonical template of the placenta. Volumetric regularization, including Jacobian-determinant and biharmonic penalties, suppresses local folding and promotes globally coherent deformations. In the motivating application to placenta MRI, our formulation jointly reconstructs individual placentas, aligns them to a population-derived implicit template, and enables voxel-wise intensity mapping in a unified canonical space. Experiments on in-vivo placenta MRI scans demonstrate improved geometric fidelity and volumetric alignment over surface-based implicit baseline methods, yielding anatomically interpretable and topologically consistent flattening suitable for group analysis.

Benjamin Billot, Neel Dey, Daniel Moyer et al. · mit

Rigid motion tracking is paramount in many medical imaging applications where movements need to be detected, corrected, or accounted for. Modern strategies rely on convolutional neural networks (CNN) and pose this problem as rigid registration. Yet, CNNs do not exploit natural symmetries in this task, as they are equivariant to translations (their outputs shift with their inputs) but not to rotations. Here we propose EquiTrack, the first method that uses recent steerable SE(3)-equivariant CNNs (E-CNN) for motion tracking. While steerable E-CNNs can extract corresponding features across different poses, testing them on noisy medical images reveals that they do not have enough learning capacity to learn noise invariance. Thus, we introduce a hybrid architecture that pairs a denoiser with an E-CNN to decouple the processing of anatomically irrelevant intensity features from the extraction of equivariant spatial features. Rigid transforms are then estimated in closed-form. EquiTrack outperforms state-of-the-art learning and optimisation methods for motion tracking in adult brain MRI and fetal MRI time series. Our code is available at https://github.com/BBillot/EquiTrack.

Yingcheng Liu, Athena Taymourtash, Yang Liu et al.

Automated analysis of articulated bodies is crucial in medical imaging. Existing surface-based models often ignore internal volumetric structures and rely on deformation methods that lack anatomical consistency guarantees. To address this problem, we introduce a differentiable volumetric body model based on the Skinned Multi-Person Linear (SMPL) formulation, driven by a new Kinematic Tree-based Log-Euclidean PolyRigid (KTPolyRigid) transform. KTPolyRigid resolves Lie algebra ambiguities associated with large, non-local articulated motions, and encourages smooth, bijective volumetric mappings. Evaluated on 53 fetal MRI volumes, KTPolyRigid yields deformation fields with significantly fewer folding artifacts. Furthermore, our framework enables robust groupwise image registration and a label-efficient, template-based segmentation of fetal organs. It provides a robust foundation for standardized volumetric analysis of articulated bodies in medical imaging.

Vivek Gopalakrishnan, Neel Dey, David-Dimitris Chlorogiannis et al. · mit

The integration of artificial intelligence in image-guided interventions holds transformative potential, promising to extract 3D geometric and quantitative information from conventional 2D imaging modalities during complex procedures. Achieving this requires the rapid and precise alignment of 2D intraoperative images (e.g., X-ray) with 3D preoperative volumes (e.g., CT, MRI). However, current 2D/3D registration methods fail across the broad spectrum of procedures dependent on X-ray guidance: traditional optimization techniques require custom parameter tuning for each subject, whereas neural networks trained on small datasets do not generalize to new patients or require labor-intensive manual annotations, increasing clinical burden and precluding application to new anatomical targets. To address these challenges, we present xvr, a fully automated framework for training patient-specific neural networks for 2D/3D registration. xvr uses physics-based simulation to generate abundant high-quality training data from a patient's own preoperative volumetric imaging, thereby overcoming the inherently limited ability of supervised models to generalize to new patients and procedures. Furthermore, xvr requires only 5 minutes of training per patient, making it suitable for emergency interventions as well as planned procedures. We perform the largest evaluation of a 2D/3D registration algorithm on real X-ray data to date and find that xvr robustly generalizes across a diverse dataset comprising multiple anatomical structures, imaging modalities, and hospitals. Across surgical tasks, xvr achieves submillimeter-accurate registration at intraoperative speeds, improving upon existing methods by an order of magnitude. xvr is released as open-source software freely available at https://github.com/eigenvivek/xvr.

Sean I. Young, Yaël Balbastre, Bruce Fischl et al. · harvard

In magnetic resonance imaging (MRI), slice-to-volume reconstruction (SVR) refers to computational reconstruction of an unknown 3D magnetic resonance volume from stacks of 2D slices corrupted by motion. While promising, current SVR methods require multiple slice stacks for accurate 3D reconstruction, leading to long scans and limiting their use in time-sensitive applications such as fetal fMRI. Here, we propose a SVR method that overcomes the shortcomings of previous work and produces state-of-the-art reconstructions in the presence of extreme inter-slice motion. Inspired by the recent success of single-view depth estimation methods, we formulate SVR as a single-stack motion estimation task and train a fully convolutional network to predict a motion stack for a given slice stack, producing a 3D reconstruction as a byproduct of the predicted motion. Extensive experiments on the SVR of adult and fetal brains demonstrate that our fully convolutional method is twice as accurate as previous SVR methods. Our code is available at github.com/seannz/svr.

Mohammadhossein Momeni, Vivek Gopalakrishnan, Neel Dey et al. · mit

We present DiffVox, a self-supervised framework for Cone-Beam Computed Tomography (CBCT) reconstruction by directly optimizing a voxelgrid representation using physics-based differentiable X-ray rendering. Further, we investigate how the different implementations of the X-ray image formation model in the renderer affect the quality of 3D reconstruction and novel view synthesis. When combined with our regularized voxel-based learning framework, we find that using an exact implementation of the discrete Beer-Lambert law for X-ray attenuation in the renderer outperforms both widely used iterative CBCT reconstruction algorithms and modern neural field approaches, particularly when given only a few input views. As a result, we reconstruct high-fidelity 3D CBCT volumes from fewer X-rays, potentially reducing ionizing radiation exposure and improving diagnostic utility. Our implementation is available at https://github.com/hossein-momeni/DiffVox.

S. Mazdak Abulnaga, Neel Dey, Sean I. Young et al. · mit

Blood oxygen level dependent (BOLD) MRI time series with maternal hyperoxia can assess placental oxygenation and function. Measuring precise BOLD changes in the placenta requires accurate temporal placental segmentation and is confounded by fetal and maternal motion, contractions, and hyperoxia-induced intensity changes. Current BOLD placenta segmentation methods warp a manually annotated subject-specific template to the entire time series. However, as the placenta is a thin, elongated, and highly non-rigid organ subject to large deformations and obfuscated edges, existing work cannot accurately segment the placental shape, especially near boundaries. In this work, we propose a machine learning segmentation framework for placental BOLD MRI and apply it to segmenting each volume in a time series. We use a placental-boundary weighted loss formulation and perform a comprehensive evaluation across several popular segmentation objectives. Our model is trained and tested on a cohort of 91 subjects containing healthy fetuses, fetuses with fetal growth restriction, and mothers with high BMI. Biomedically, our model performs reliably in segmenting volumes in both normoxic and hyperoxic points in the BOLD time series. We further find that boundary-weighting increases placental segmentation performance by 8.3% and 6.0% Dice coefficient for the cross-entropy and signed distance transform objectives, respectively. Our code and trained model is available at https://github.com/mabulnaga/automatic-placenta-segmentation.

Jian Wang, Razieh Faghihpirayesh, Polina Golland et al.

In this paper, we introduce SpaER, a pioneering method for fetal motion tracking that leverages equivariant filters and self-attention mechanisms to effectively learn spatio-temporal representations. Different from conventional approaches that statically estimate fetal brain motions from pairs of images, our method dynamically tracks the rigid movement patterns of the fetal head across temporal and spatial dimensions. Specifically, we first develop an equivariant neural network that efficiently learns rigid motion sequences through low-dimensional spatial representations of images. Subsequently, we learn spatio-temporal representations by incorporating time encoding and self-attention neural network layers. This approach allows for the capture of long-term dependencies of fetal brain motion and addresses alignment errors due to contrast changes and severe motion artifacts. Our model also provides a geometric deformation estimation that properly addresses image distortions among all time frames. To the best of our knowledge, our approach is the first to learn spatial-temporal representations via deep neural networks for fetal motion tracking without data augmentation. We validated our model using real fetal echo-planar images with simulated and real motions. Our method carries significant potential value in accurately measuring, tracking, and correcting fetal motion in fetal MRI sequences.

Benjamin Billot, Ramya Muthukrishnan, Esra Abaci-Turk et al.

Unsupervised registration strategies bypass requirements in ground truth transforms or segmentations by optimising similarity metrics between fixed and moved volumes. Among these methods, a recent subclass of approaches based on unsupervised keypoint detection stand out as very promising for interpretability. Specifically, these methods train a network to predict feature maps for fixed and moving images, from which explainable centres of mass are computed to obtain point clouds, that are then aligned in closed-form. However, the features returned by the network often yield spatially diffuse patterns that are hard to interpret, thus undermining the purpose of keypoint-based registration. Here, we propose a three-fold loss to regularise the spatial distribution of the features. First, we use the KL divergence to model features as point spread functions that we interpret as probabilistic keypoints. Then, we sharpen the spatial distributions of these features to increase the precision of the detected landmarks. Finally, we introduce a new repulsive loss across keypoints to encourage spatial diversity. Overall, our loss considerably improves the interpretability of the features, which now correspond to precise and anatomically meaningful landmarks. We demonstrate our three-fold loss in foetal rigid motion tracking and brain MRI affine registration tasks, where it not only outperforms state-of-the-art unsupervised strategies, but also bridges the gap with state-of-the-art supervised methods. Our code is available at https://github.com/BenBillot/spatial_regularisation.

Sebastian Diaz, Benjamin Billot, Neel Dey et al. · mit

Fetal motion is a critical indicator of neurological development and intrauterine health, yet its quantification remains challenging, particularly at earlier gestational ages (GA). Current methods track fetal motion by predicting the location of annotated landmarks on 3D echo planar imaging (EPI) time-series, primarily in third-trimester fetuses. The predicted landmarks enable simplification of the fetal body for downstream analysis. While these methods perform well within their training age distribution, they consistently fail to generalize to early GAs due to significant anatomical changes in both mother and fetus across gestation, as well as the difficulty of obtaining annotated early GA EPI data. In this work, we develop a cross-population data augmentation framework that enables pose estimation models to robustly generalize to younger GA clinical cohorts using only annotated images from older GA cohorts. Specifically, we introduce a fetal-specific augmentation strategy that simulates the distinct intrauterine environment and fetal positioning of early GAs. Our experiments find that cross-population augmentation yields reduced variability and significant improvements across both older GA and challenging early GA cases. By enabling more reliable pose estimation across gestation, our work potentially facilitates early clinical detection and intervention in challenging 4D fetal imaging settings. Code is available at https://github.com/sebodiaz/cross-population-pose.

Yingcheng Liu, Peiqi Wang, Sebastian Diaz et al.

Analyzing fetal body motion and shape is paramount in prenatal diagnostics and monitoring. Existing methods for fetal MRI analysis mainly rely on anatomical keypoints or volumetric body segmentations. Keypoints simplify body structure to facilitate motion analysis, but may ignore important details of full-body shape. Body segmentations capture complete shape information but complicate temporal analysis due to large non-local fetal movements. To address these limitations, we construct a 3D articulated statistical fetal body model based on the Skinned Multi-Person Linear Model (SMPL). Our algorithm iteratively estimates body pose in the image space and body shape in the canonical pose space. This approach improves robustness to MRI motion artifacts and intensity distortions, and reduces the impact of incomplete surface observations due to challenging fetal poses. We train our model on segmentations and keypoints derived from $19,816$ MRI volumes across $53$ subjects. Our model captures body shape and motion across time series and provides intuitive visualization. Furthermore, it enables automated anthropometric measurements traditionally difficult to obtain from segmentations and keypoints. When tested on unseen fetal body shapes, our method yields a surface alignment error of $3.2$ mm for $3$ mm MRI voxel size. To our knowledge, this represents the first 3D articulated statistical fetal body model, paving the way for enhanced fetal motion and shape analysis in prenatal diagnostics. The code is available at https://github.com/MedicalVisionGroup/fetal-smpl .

Sungmin Hong, Anna K. Bonkhoff, Andrew Hoopes et al.

Despite the superior performance of Deep Learning (DL) on numerous segmentation tasks, the DL-based approaches are notoriously overconfident about their prediction with highly polarized label probability. This is often not desirable for many applications with the inherent label ambiguity even in human annotations. This challenge has been addressed by leveraging multiple annotations per image and the segmentation uncertainty. However, multiple per-image annotations are often not available in a real-world application and the uncertainty does not provide full control on segmentation results to users. In this paper, we propose novel methods to improve the segmentation probability estimation without sacrificing performance in a real-world scenario that we have only one ambiguous annotation per image. We marginalize the estimated segmentation probability maps of networks that are encouraged to under-/over-segment with the varying Tversky loss without penalizing balanced segmentation. Moreover, we propose a unified hypernetwork ensemble method to alleviate the computational burden of training multiple networks. Our approaches successfully estimated the segmentation probability maps that reflected the underlying structures and provided the intuitive control on segmentation for the challenging 3D medical image segmentation. Although the main focus of our proposed methods is not to improve the binary segmentation performance, our approaches marginally outperformed the state-of-the-arts. The codes are available at \url{https://github.com/sh4174/HypernetEnsemble}.

S. Mazdak Abulnaga, Esra Abaci Turk, Mikhail Bessmeltsev et al.

We present a volumetric mesh-based algorithm for parameterizing the placenta to a flattened template to enable effective visualization of local anatomy and function. MRI shows potential as a research tool as it provides signals directly related to placental function. However, due to the curved and highly variable in vivo shape of the placenta, interpreting and visualizing these images is difficult. We address interpretation challenges by mapping the placenta so that it resembles the familiar ex vivo shape. We formulate the parameterization as an optimization problem for mapping the placental shape represented by a volumetric mesh to a flattened template. We employ the symmetric Dirichlet energy to control local distortion throughout the volume. Local injectivity in the mapping is enforced by a constrained line search during the gradient descent optimization. We validate our method using a research study of 111 placental shapes extracted from BOLD MRI images. Our mapping achieves sub-voxel accuracy in matching the template while maintaining low distortion throughout the volume. We demonstrate how the resulting flattening of the placenta improves visualization of anatomy and function. Our code is freely available at https://github.com/mabulnaga/placenta-flattening .

Sungmin Hong, Razvan Marinescu, Adrian V. Dalca et al.

Image synthesis via Generative Adversarial Networks (GANs) of three-dimensional (3D) medical images has great potential that can be extended to many medical applications, such as, image enhancement and disease progression modeling. However, current GAN technologies for 3D medical image synthesis need to be significantly improved to be readily adapted to real-world medical problems. In this paper, we extend the state-of-the-art StyleGAN2 model, which natively works with two-dimensional images, to enable 3D image synthesis. In addition to the image synthesis, we investigate the controllability and interpretability of the 3D-StyleGAN via style vectors inherited form the original StyleGAN2 that are highly suitable for medical applications: (i) the latent space projection and reconstruction of unseen real images, and (ii) style mixing. We demonstrate the 3D-StyleGAN's performance and feasibility with ~12,000 three-dimensional full brain MR T1 images, although it can be applied to any 3D volumetric images. Furthermore, we explore different configurations of hyperparameters to investigate potential improvement of the image synthesis with larger networks. The codes and pre-trained networks are available online: https://github.com/sh4174/3DStyleGAN.

Mohammad Haft-Javaherian, Martin Villiger, Kenichiro Otsuka et al.

Intravascular ultrasound and optical coherence tomography are widely available for characterizing coronary stenoses and provide critical vessel parameters to optimize percutaneous intervention. Intravascular polarization-sensitive optical coherence tomography (PS-OCT) simultaneously provides high-resolution cross-sectional images of vascular structures while also revealing preponderant tissue components such as collagen and smooth muscle and thereby enhances plaque characterization. Automated interpretation of these features promises to facilitate the objective clinical investigation of the natural history and significance of coronary atheromas. Here, we propose a convolutional neural network model, optimized using a new multi-term loss function, to classify the lumen, intima, and media layers in addition to the guidewire and plaque shadows. We demonstrate that our multi-class classification model outperforms state-of-the-art methods in detecting the coronary anatomical layers. Furthermore, the proposed model segments two classes of common imaging artifacts and detects the anatomical layers within the thickened vessel wall regions that were excluded from analysis by other studies. The source code and the trained model are publicly available at https://github.com/mhaft/OCTseg

Juan Eugenio Iglesias, Benjamin Billot, Yael Balbastre et al.

Most existing algorithms for automatic 3D morphometry of human brain MRI scans are designed for data with near-isotropic voxels at approximately 1 mm resolution, and frequently have contrast constraints as well - typically requiring T1 scans (e.g., MP-RAGE). This limitation prevents the analysis of millions of MRI scans acquired with large inter-slice spacing ("thick slice") in clinical settings every year. The inability to quantitatively analyze these scans hinders the adoption of quantitative neuroimaging in healthcare, and precludes research studies that could attain huge sample sizes and hence greatly improve our understanding of the human brain. Recent advances in CNNs are producing outstanding results in super-resolution and contrast synthesis of MRI. However, these approaches are very sensitive to the contrast, resolution and orientation of the input images, and thus do not generalize to diverse clinical acquisition protocols - even within sites. Here we present SynthSR, a method to train a CNN that receives one or more thick-slice scans with different contrast, resolution and orientation, and produces an isotropic scan of canonical contrast (typically a 1 mm MP-RAGE). The presented method does not require any preprocessing, e.g., skull stripping or bias field correction. Crucially, SynthSR trains on synthetic input images generated from 3D segmentations, and can thus be used to train CNNs for any combination of contrasts, resolutions and orientations without high-resolution training data. We test the images generated with SynthSR in an array of common downstream analyses, and show that they can be reliably used for subcortical segmentation and volumetry, image registration (e.g., for tensor-based morphometry), and, if some image quality requirements are met, even cortical thickness morphometry. The source code is publicly available at github.com/BBillot/SynthSR.

Adrian V. Dalca, Ramesh Sridharan, Mert R. Sabuncu et al.

We present a semi-parametric generative model for predicting anatomy of a patient in subsequent scans following a single baseline image. Such predictive modeling promises to facilitate novel analyses in both voxel-level studies and longitudinal biomarker evaluation. We capture anatomical change through a combination of population-wide regression and a non-parametric model of the subject's health based on individual genetic and clinical indicators. In contrast to classical correlation and longitudinal analysis, we focus on predicting new observations from a single subject observation. We demonstrate prediction of follow-up anatomical scans in the ADNI cohort, and illustrate a novel analysis approach that compares a patient's scans to the predicted subject-specific healthy anatomical trajectory. The code is available at https://github.com/adalca/voxelorb.

Ruizhi Liao, Daniel Moyer, Polina Golland et al.

Estimating mutual information between continuous random variables is often intractable and extremely challenging for high-dimensional data. Recent progress has leveraged neural networks to optimize variational lower bounds on mutual information. Although showing promise for this difficult problem, the variational methods have been theoretically and empirically proven to have serious statistical limitations: 1) many methods struggle to produce accurate estimates when the underlying mutual information is either low or high; 2) the resulting estimators may suffer from high variance. Our approach is based on training a classifier that provides the probability that a data sample pair is drawn from the joint distribution rather than from the product of its marginal distributions. Moreover, we establish a direct connection between mutual information and the average log odds estimate produced by the classifier on a test set, leading to a simple and accurate estimator of mutual information. We show theoretically that our method and other variational approaches are equivalent when they achieve their optimum, while our method sidesteps the variational bound. Empirical results demonstrate high accuracy of our approach and the advantages of our estimator in the context of representation learning. Our demo is available at https://github.com/RayRuizhiLiao/demi_mi_estimator.

Geeticka Chauhan, Ruizhi Liao, William Wells et al.

We propose and demonstrate a novel machine learning algorithm that assesses pulmonary edema severity from chest radiographs. While large publicly available datasets of chest radiographs and free-text radiology reports exist, only limited numerical edema severity labels can be extracted from radiology reports. This is a significant challenge in learning such models for image classification. To take advantage of the rich information present in the radiology reports, we develop a neural network model that is trained on both images and free-text to assess pulmonary edema severity from chest radiographs at inference time. Our experimental results suggest that the joint image-text representation learning improves the performance of pulmonary edema assessment compared to a supervised model trained on images only. We also show the use of the text for explaining the image classification by the joint model. To the best of our knowledge, our approach is the first to leverage free-text radiology reports for improving the image model performance in this application. Our code is available at https://github.com/RayRuizhiLiao/joint_chestxray.

Nalini M. Singh, Juan Eugenio Iglesias, Elfar Adalsteinsson et al.

We propose neural network layers that explicitly combine frequency and image feature representations and show that they can be used as a versatile building block for reconstruction from frequency space data. Our work is motivated by the challenges arising in MRI acquisition where the signal is a corrupted Fourier transform of the desired image. The proposed joint learning schemes enable both correction of artifacts native to the frequency space and manipulation of image space representations to reconstruct coherent image structures at every layer of the network. This is in contrast to most current deep learning approaches for image reconstruction that treat frequency and image space features separately and often operate exclusively in one of the two spaces. We demonstrate the advantages of joint convolutional learning for a variety of tasks, including motion correction, denoising, reconstruction from undersampled acquisitions, and combined undersampling and motion correction on simulated and real world multicoil MRI data. The joint models produce consistently high quality output images across all tasks and datasets. When integrated into a state of the art unrolled optimization network with physics-inspired data consistency constraints for undersampled reconstruction, the proposed architectures significantly improve the optimization landscape, which yields an order of magnitude reduction of training time. This result suggests that joint representations are particularly well suited for MRI signals in deep learning networks. Our code and pretrained models are publicly available at https://github.com/nalinimsingh/interlacer.

S. Mazdak Abulnaga, Esra Abaci Turk, Mikhail Bessmeltsev et al.

We present a volumetric mesh-based algorithm for flattening the placenta to a canonical template to enable effective visualization of local anatomy and function. Monitoring placental function in vivo promises to support pregnancy assessment and to improve care outcomes. We aim to alleviate visualization and interpretation challenges presented by the shape of the placenta when it is attached to the curved uterine wall. To do so, we flatten the volumetric mesh that captures placental shape to resemble the well-studied ex vivo shape. We formulate our method as a map from the in vivo shape to a flattened template that minimizes the symmetric Dirichlet energy to control distortion throughout the volume. Local injectivity is enforced via constrained line search during gradient descent. We evaluate the proposed method on 28 placenta shapes extracted from MRI images in a clinical study of placental function. We achieve sub-voxel accuracy in mapping the boundary of the placenta to the template while successfully controlling distortion throughout the volume. We illustrate how the resulting mapping of the placenta enhances visualization of placental anatomy and function. Our code is freely available at https://github.com/mabulnaga/placenta-flattening .

Razvan V. Marinescu, Marco Lorenzi, Stefano B. Blumberg et al.

We introduce Disease Knowledge Transfer (DKT), a novel technique for transferring biomarker information between related neurodegenerative diseases. DKT infers robust multimodal biomarker trajectories in rare neurodegenerative diseases even when only limited, unimodal data is available, by transferring information from larger multimodal datasets from common neurodegenerative diseases. DKT is a joint-disease generative model of biomarker progressions, which exploits biomarker relationships that are shared across diseases. Our proposed method allows, for the first time, the estimation of plausible, multimodal biomarker trajectories in Posterior Cortical Atrophy (PCA), a rare neurodegenerative disease where only unimodal MRI data is available. For this we train DKT on a combined dataset containing subjects with two distinct diseases and sizes of data available: 1) a larger, multimodal typical AD (tAD) dataset from the TADPOLE Challenge, and 2) a smaller unimodal Posterior Cortical Atrophy (PCA) dataset from the Dementia Research Centre (DRC), for which only a limited number of Magnetic Resonance Imaging (MRI) scans are available. Although validation is challenging due to lack of data in PCA, we validate DKT on synthetic data and two patient datasets (TADPOLE and PCA cohorts), showing it can estimate the ground truth parameters in the simulation and predict unseen biomarkers on the two patient datasets. While we demonstrated DKT on Alzheimer's variants, we note DKT is generalisable to other forms of related neurodegenerative diseases. Source code for DKT is available online: https://github.com/mrazvan22/dkt.

Adrian V. Dalca, Katherine L. Bouman, William T. Freeman et al.

We present an algorithm for creating high resolution anatomically plausible images consistent with acquired clinical brain MRI scans with large inter-slice spacing. Although large data sets of clinical images contain a wealth of information, time constraints during acquisition result in sparse scans that fail to capture much of the anatomy. These characteristics often render computational analysis impractical as many image analysis algorithms tend to fail when applied to such images. Highly specialized algorithms that explicitly handle sparse slice spacing do not generalize well across problem domains. In contrast, we aim to enable application of existing algorithms that were originally developed for high resolution research scans to significantly undersampled scans. We introduce a generative model that captures fine-scale anatomical structure across subjects in clinical image collections and derive an algorithm for filling in the missing data in scans with large inter-slice spacing. Our experimental results demonstrate that the resulting method outperforms state-of-the-art upsampling super-resolution techniques, and promises to facilitate subsequent analysis not previously possible with scans of this quality. Our implementation is freely available at https://github.com/adalca/papago .

Peiqi Wang, Yikang Shen, Zhen Guo et al.

We aim to select data subsets for the fine-tuning of large language models to more effectively follow instructions. Prior work has emphasized the importance of diversity in dataset curation but relied on heuristics such as the number of tasks. In this paper, we use determinantal point processes to capture the diversity and quality of instruction tuning datasets for subset selection. We propose to measure dataset diversity with log determinant distance that is the distance between the dataset of interest and a maximally diverse reference dataset. Our experiments demonstrate that the proposed diversity measure in the normalized weight gradient space is correlated with downstream instruction-following performance. Consequently, it can be used to inform when data selection is the most helpful and to analyze dataset curation strategies. We demonstrate the utility of our approach on various instruction tuning datasets.

Neel Dey, Benjamin Billot, Hallee E. Wong et al. · mit

Current volumetric biomedical foundation models struggle to generalize as public 3D datasets are small and do not cover the broad diversity of medical procedures, conditions, anatomical regions, and imaging protocols. We address this by creating a representation learning method that instead anticipates strong domain shifts at training time itself. We first propose a data engine that synthesizes highly variable training samples that would enable generalization to new biomedical contexts. To then train a single 3D network for any voxel-level task, we develop a contrastive learning method that pretrains the network to be stable against nuisance imaging variation simulated by the data engine, a key inductive bias for generalization. This network's features can be used as robust representations of input images for downstream tasks and its weights provide a strong, dataset-agnostic initialization for finetuning on new datasets. As a result, we set new standards across both multimodality registration and few-shot segmentation, a first for any 3D biomedical vision model, all without (pre-)training on any existing dataset of real images.

Molin Zhang, Polina Golland, Patricia Ellen Grant et al.

The quality of fetal MRI is significantly affected by unpredictable and substantial fetal motion, leading to the introduction of artifacts even when fast acquisition sequences are employed. The development of 3D real-time fetal pose estimation approaches on volumetric EPI fetal MRI opens up a promising avenue for fetal motion monitoring and prediction. Challenges arise in fetal pose estimation due to limited number of real scanned fetal MR training images, hindering model generalization when the acquired fetal MRI lacks adequate pose. In this study, we introduce FetalDiffusion, a novel approach utilizing a conditional diffusion model to generate 3D synthetic fetal MRI with controllable pose. Additionally, an auxiliary pose-level loss is adopted to enhance model performance. Our work demonstrates the success of this proposed model by producing high-quality synthetic fetal MRI images with accurate and recognizable fetal poses, comparing favorably with in-vivo real fetal MRI. Furthermore, we show that the integration of synthetic fetal MR images enhances the fetal pose estimation model's performance, particularly when the number of available real scanned data is limited resulting in 15.4% increase in PCK and 50.2% reduced in mean error. All experiments are done on a single 32GB V100 GPU. Our method holds promise for improving real-time tracking models, thereby addressing fetal motion issues more effectively.

Reuben Dorent, Polina Golland, William Wells

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.