-
Topologically faithful multi-class segmentation in medical images
Authors:
Alexander H. Berger,
Nico Stucki,
Laurin Lux,
Vincent Buergin,
Suprosanna Shit,
Anna Banaszak,
Daniel Rueckert,
Ulrich Bauer,
Johannes C. Paetzold
Abstract:
Topological accuracy in medical image segmentation is a highly important property for downstream applications such as network analysis and flow modeling in vessels or cell counting. Recently, significant methodological advancements have brought well-founded concepts from algebraic topology to binary segmentation. However, these approaches have been underexplored in multi-class segmentation scenari…
▽ More
Topological accuracy in medical image segmentation is a highly important property for downstream applications such as network analysis and flow modeling in vessels or cell counting. Recently, significant methodological advancements have brought well-founded concepts from algebraic topology to binary segmentation. However, these approaches have been underexplored in multi-class segmentation scenarios, where topological errors are common. We propose a general loss function for topologically faithful multi-class segmentation extending the recent Betti matching concept, which is based on induced matchings of persistence barcodes. We project the N-class segmentation problem to N single-class segmentation tasks, which allows us to use 1-parameter persistent homology making training of neural networks computationally feasible. We validate our method on a comprehensive set of four medical datasets with highly variant topological characteristics. Our loss formulation significantly enhances topological correctness in cardiac, cell, artery-vein, and Circle of Willis segmentation.
△ Less
Submitted 16 March, 2024;
originally announced March 2024.
-
On the Localization of Ultrasound Image Slices within Point Distribution Models
Authors:
Lennart Bastian,
Vincent Bürgin,
Ha Young Kim,
Alexander Baumann,
Benjamin Busam,
Mahdi Saleh,
Nassir Navab
Abstract:
Thyroid disorders are most commonly diagnosed using high-resolution Ultrasound (US). Longitudinal nodule tracking is a pivotal diagnostic protocol for monitoring changes in pathological thyroid morphology. This task, however, imposes a substantial cognitive load on clinicians due to the inherent challenge of maintaining a mental 3D reconstruction of the organ. We thus present a framework for autom…
▽ More
Thyroid disorders are most commonly diagnosed using high-resolution Ultrasound (US). Longitudinal nodule tracking is a pivotal diagnostic protocol for monitoring changes in pathological thyroid morphology. This task, however, imposes a substantial cognitive load on clinicians due to the inherent challenge of maintaining a mental 3D reconstruction of the organ. We thus present a framework for automated US image slice localization within a 3D shape representation to ease how such sonographic diagnoses are carried out. Our proposed method learns a common latent embedding space between US image patches and the 3D surface of an individual's thyroid shape, or a statistical aggregation in the form of a statistical shape model (SSM), via contrastive metric learning. Using cross-modality registration and Procrustes analysis, we leverage features from our model to register US slices to a 3D mesh representation of the thyroid shape. We demonstrate that our multi-modal registration framework can localize images on the 3D surface topology of a patient-specific organ and the mean shape of an SSM. Experimental results indicate slice positions can be predicted within an average of 1.2 mm of the ground-truth slice location on the patient-specific 3D anatomy and 4.6 mm on the SSM, exemplifying its usefulness for slice localization during sonographic acquisitions. Code is publically available: \href{https://github.com/vuenc/slice-to-shape}{https://github.com/vuenc/slice-to-shape}
△ Less
Submitted 1 September, 2023;
originally announced September 2023.
-
Robust vertebra identification using simultaneous node and edge predicting Graph Neural Networks
Authors:
Vincent Bürgin,
Raphael Prevost,
Marijn F. Stollenga
Abstract:
Automatic vertebra localization and identification in CT scans is important for numerous clinical applications. Much progress has been made on this topic, but it mostly targets positional localization of vertebrae, ignoring their orientation. Additionally, most methods employ heuristics in their pipeline that can be sensitive in real clinical images which tend to contain abnormalities. We introduc…
▽ More
Automatic vertebra localization and identification in CT scans is important for numerous clinical applications. Much progress has been made on this topic, but it mostly targets positional localization of vertebrae, ignoring their orientation. Additionally, most methods employ heuristics in their pipeline that can be sensitive in real clinical images which tend to contain abnormalities. We introduce a simple pipeline that employs a standard prediction with a U-Net, followed by a single graph neural network to associate and classify vertebrae with full orientation. To test our method, we introduce a new vertebra dataset that also contains pedicle detections that are associated with vertebra bodies, creating a more challenging landmark prediction, association and classification task. Our method is able to accurately associate the correct body and pedicle landmarks, ignore false positives and classify vertebrae in a simple, fully trainable pipeline avoiding application-specific heuristics. We show our method outperforms traditional approaches such as Hungarian Matching and Hidden Markov Models. We also show competitive performance on the standard VerSe challenge body identification task.
△ Less
Submitted 27 July, 2023;
originally announced August 2023.
-
S3M: Scalable Statistical Shape Modeling through Unsupervised Correspondences
Authors:
Lennart Bastian,
Alexander Baumann,
Emily Hoppe,
Vincent Bürgin,
Ha Young Kim,
Mahdi Saleh,
Benjamin Busam,
Nassir Navab
Abstract:
Statistical shape models (SSMs) are an established way to represent the anatomy of a population with various clinically relevant applications. However, they typically require domain expertise, and labor-intensive landmark annotations to construct. We address these shortcomings by proposing an unsupervised method that leverages deep geometric features and functional correspondences to simultaneousl…
▽ More
Statistical shape models (SSMs) are an established way to represent the anatomy of a population with various clinically relevant applications. However, they typically require domain expertise, and labor-intensive landmark annotations to construct. We address these shortcomings by proposing an unsupervised method that leverages deep geometric features and functional correspondences to simultaneously learn local and global shape structures across population anatomies. Our pipeline significantly improves unsupervised correspondence estimation for SSMs compared to baseline methods, even on highly irregular surface topologies. We demonstrate this for two different anatomical structures: the thyroid and a multi-chamber heart dataset. Furthermore, our method is robust enough to learn from noisy neural network predictions, potentially enabling scaling SSMs to larger patient populations without manual segmentation annotation.
△ Less
Submitted 24 July, 2023; v1 submitted 15 April, 2023;
originally announced April 2023.
-
On Game Theory Using Stochastic Tail Orders
Authors:
Stefan Rass,
Sandra König,
Stefan Schauer,
Vincent Bürgin,
Jeremias Epperlein,
Fabian Wirth
Abstract:
We consider a family of distributions on which natural tail orders can be constructed upon a representation of a distribution by a (single) hyper-real number. Past research revealed that the ordering can herein strongly depend on the particular model of the hyperreals, specifically the underlying ultrafilter. Hence, our distribution family is constructed to order invariantly of an ultrafilter. Mor…
▽ More
We consider a family of distributions on which natural tail orders can be constructed upon a representation of a distribution by a (single) hyper-real number. Past research revealed that the ordering can herein strongly depend on the particular model of the hyperreals, specifically the underlying ultrafilter. Hence, our distribution family is constructed to order invariantly of an ultrafilter. Moreover, we prove that it lies dense in the set of all distributions with the (same) compact support, w.r.t. the supremum norm. Overall, this work resents a correction to [10, 12], in response to recent findings of [2].
△ Less
Submitted 19 April, 2022; v1 submitted 2 August, 2021;
originally announced August 2021.
-
Remarks on the tail order on moment sequences
Authors:
Vincent Bürgin,
Jeremias Epperlein,
Fabian Wirth
Abstract:
We consider positively supported Borel measures for which all moments exist. On the set of compactly supported measures in this class a partial order is defined via eventual dominance of the moment sequences. Special classes are identified on which the order is total, but it is shown that already for the set of distributions with compactly supported smooth densities the order is not total. In part…
▽ More
We consider positively supported Borel measures for which all moments exist. On the set of compactly supported measures in this class a partial order is defined via eventual dominance of the moment sequences. Special classes are identified on which the order is total, but it is shown that already for the set of distributions with compactly supported smooth densities the order is not total. In particular we construct a pair of measures with smooth density for which infinitely many moments agree and another one for which the moments alternate infinitely often. This disproves some recently published claims to the contrary. Some consequences for games with distributional payoffs are discussed.
△ Less
Submitted 21 March, 2022; v1 submitted 21 April, 2021;
originally announced April 2021.
-
Distribution-Valued Games -- Overview, Analysis, and a Segmentation-Based Approach
Authors:
Vincent Bürgin
Abstract:
The paper [Ras15a] introduced distribution-valued games. This game-theoretic model uses probability distributions as payoffs for games in order to express uncertainty about the payoffs. The player's preferences for different payoffs are expressed by a stochastic order which we call the tail order.
This thesis formalizes distribution-valued games with preferences expressed by general stochastic o…
▽ More
The paper [Ras15a] introduced distribution-valued games. This game-theoretic model uses probability distributions as payoffs for games in order to express uncertainty about the payoffs. The player's preferences for different payoffs are expressed by a stochastic order which we call the tail order.
This thesis formalizes distribution-valued games with preferences expressed by general stochastic orders, and specifically analyzes properties of the tail order. It identifies sufficient conditions for tail-order preference to hold, but also finds that some claims in [Ras15a] about the tail order are incorrect, for which counter-examples are constructed. In particular, it is demonstrated that a proof for the totality of the order on a certain set of distributions contains an error; the thesis proceeds to show that the ordering is not total on the slightly less restricted set of distributions with non-negative bounded support. It is also shown that not all tail-ordered games have mixed-strategy Nash equilibria, and in fact almost all tail-ordered games with finitely-supported payoff distributions can only have a Nash equilibrium if they have a pure-strategy Nash equilibrium.
The thesis subsequently extends an idea from [AM19] and proposes a new solution concept for distribution-valued games. This concept is based on constructing multi-objective real-valued games from distribution-valued games by segmenting their payoff distributions.
△ Less
Submitted 19 March, 2021;
originally announced March 2021.
