Mariusz Bujny
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mariusz Bujny has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Fabian Duddeck
School of Engineering and Design, Technical University of Munich, Munich, Germany
11 shared publications Nikola Aulig
Honda Research Institute Europe GmbH, Offenbach/Main, Germany
6 shared publications- Jakub Nalepa
Silesian University of Technology
5 shared publications - Marcin Kostur
Graylight Imaging
5 shared publications - Markus Olhofer
Honda Research Institute Europe GmbH
5 shared publications - Karol Miszalski-Jamka
Graylight Imaging
4 shared publications
Latest Publications
- 1 February 2026
- Published by Springer Nature
- p. 563-585
Optimization Is Not Enough: Why Problem Formulation Deserves Equal Attention
- 1 February 2026
- Published by Springer Nature
- p. 563-585
Optimization Is Not Enough: Why Problem Formulation Deserves Equal Attention
- Top Cited Papers
- 2 January 2026
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 202, 111421
Coronary artery segmentation in non-contrast calcium scoring CT images using deep learning
- 2 January 2026
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 202, 111421
Top Cited PapersCoronary artery segmentation in non-contrast calcium scoring CT images using deep learning
- Open Access
Enhancing Coronary Artery Calcium Scoring via Multi-Organ Segmentation on Non-Contrast Cardiac Computed Tomography
Despite coronary artery calcium scoring being considered a largely solved problem within the realm of medical artificial intelligence, this paper argues that significant improvements can still be made. By shifting the focus from pathology detection to a deeper understanding of anatomy, the novel algorithm proposed in the paper both achieves high accuracy in coronary artery calcium scoring and offers enhanced interpretability of the results. This approach not only aids in the precise quantification of calcifications in coronary arteries, but also provides valuable insights into the underlying anatomical structures. Through this anatomically-informed methodology, the paper shows how a nuanced understanding of the heart's anatomy can lead to more accurate and interpretable results in the field of cardiovascular health. We demonstrate the superior accuracy of the proposed method by evaluating it on an open-source multi-vendor dataset, where we obtain results at the inter-observer level, surpassing the current state of the art. Finally, the qualitative analyses show the practical value of the algorithm in such tasks as labeling coronary artery calcifications, identifying aortic calcifications, and filtering out false positive detections due to noise.
Open AccessEnhancing Coronary Artery Calcium Scoring via Multi-Organ Segmentation on Non-Contrast Cardiac Computed Tomography
Despite coronary artery calcium scoring being considered a largely solved problem within the realm of medical artificial intelligence, this paper argues that significant improvements can still be made. By shifting the focus from pathology detection to a deeper understanding of anatomy, the novel algorithm proposed in the paper both achieves high accuracy in coronary artery calcium scoring and offers enhanced interpretability of the results. This approach not only aids in the precise quantification of calcifications in coronary arteries, but also provides valuable insights into the underlying anatomical structures. Through this anatomically-informed methodology, the paper shows how a nuanced understanding of the heart's anatomy can lead to more accurate and interpretable results in the field of cardiovascular health. We demonstrate the superior accuracy of the proposed method by evaluating it on an open-source multi-vendor dataset, where we obtain results at the inter-observer level, surpassing the current state of the art. Finally, the qualitative analyses show the practical value of the algorithm in such tasks as labeling coronary artery calcifications, identifying aortic calcifications, and filtering out false positive detections due to noise.
- 3 October 2024
- Published by Springer Nature
- p. 572-581
Seeing the Invisible: On Aortic Valve Reconstruction in Non-contrast CT
- 3 October 2024
- Published by Springer Nature
- p. 572-581
Seeing the Invisible: On Aortic Valve Reconstruction in Non-contrast CT
- Open Access
Coronary artery segmentation in non-contrast calcium scoring CT images using deep learning
Precise localization of coronary arteries in Computed Tomography (CT) scans is critical from the perspective of medical assessment of coronary artery disease. Although various methods exist that offer high-quality segmentation of coronary arteries in cardiac contrast-enhanced CT scans, the potential of less invasive, non-contrast CT in this area is still not fully exploited. Since such fine anatomical structures are hardly visible in this type of medical images, the existing methods are characterized by high recall and low precision, and are used mainly for filtering of atherosclerotic plaques in the context of calcium scoring. In this paper, we address this research gap and introduce a deep learning algorithm for segmenting coronary arteries in multi-vendor ECG-gated non-contrast cardiac CT images which benefits from a novel framework for semi-automatic generation of Ground Truth (GT) via image registration. We hypothesize that the proposed GT generation process is much more efficient in this case than manual segmentation, since it allows for a fast generation of large volumes of diverse data, which leads to well-generalizing models. To investigate and thoroughly evaluate the segmentation quality based on such an approach, we propose a novel method for manual mesh-to-image registration, which is used to create our test-GT. The experimental study shows that the trained model has significantly higher accuracy than the GT used for training, and leads to the Dice and clDice metrics close to the interrater variability.
Open AccessCoronary artery segmentation in non-contrast calcium scoring CT images using deep learning
Precise localization of coronary arteries in Computed Tomography (CT) scans is critical from the perspective of medical assessment of coronary artery disease. Although various methods exist that offer high-quality segmentation of coronary arteries in cardiac contrast-enhanced CT scans, the potential of less invasive, non-contrast CT in this area is still not fully exploited. Since such fine anatomical structures are hardly visible in this type of medical images, the existing methods are characterized by high recall and low precision, and are used mainly for filtering of atherosclerotic plaques in the context of calcium scoring. In this paper, we address this research gap and introduce a deep learning algorithm for segmenting coronary arteries in multi-vendor ECG-gated non-contrast cardiac CT images which benefits from a novel framework for semi-automatic generation of Ground Truth (GT) via image registration. We hypothesize that the proposed GT generation process is much more efficient in this case than manual segmentation, since it allows for a fast generation of large volumes of diverse data, which leads to well-generalizing models. To investigate and thoroughly evaluate the segmentation quality based on such an approach, we propose a novel method for manual mesh-to-image registration, which is used to create our test-GT. The experimental study shows that the trained model has significantly higher accuracy than the GT used for training, and leads to the Dice and clDice metrics close to the interrater variability.
- 6 August 2022
- Published by Springer Nature
- p. 104-114
Clustering Topologically-Optimized Designs Based on Structural Deformation
- 6 August 2022
- Published by Springer Nature
- p. 104-114
Clustering Topologically-Optimized Designs Based on Structural Deformation
- Open Access
- 22 April 2022
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (1) , 904-929
Evaluation of geometric similarity metrics for structural clusters generated using topology optimization
In the early stages of engineering design, multitudes of feasible designs can be generated using structural optimization methods by varying the design requirements or user preferences for different performance objectives. Data mining such potentially large datasets is a challenging task. An unsupervised data-centric approach for exploring designs is to find clusters of similar designs and recommend only the cluster representatives for review. Design similarity can be defined not only on a purely functional level but also based on geometric properties, such as size, shape, and topology. While metrics such as chamfer distance measure the geometrical differences intuitively, it is more useful for design exploration to use metrics based on geometric features, which are extracted from high-dimensional 3D geometric data using dimensionality reduction techniques. If the Euclidean distance in the geometric features is meaningful, the features can be combined with performance attributes resulting in an aggregate feature vector that can potentially be useful in design exploration based on both geometry and performance. We propose a novel approach to evaluate such derived metrics by measuring their similarity with the metrics commonly used in 3D object classification. Furthermore, we measure clustering accuracy, which is a state-of-the-art unsupervised approach to evaluate metrics. For this purpose, we use a labeled, synthetic dataset with topologically complex designs. From our results, we conclude that Pointcloud Autoencoder is promising in encoding geometric features and developing a comprehensive design exploration method.
- 22 April 2022
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (1) , 904-929
Open AccessEvaluation of geometric similarity metrics for structural clusters generated using topology optimization
In the early stages of engineering design, multitudes of feasible designs can be generated using structural optimization methods by varying the design requirements or user preferences for different performance objectives. Data mining such potentially large datasets is a challenging task. An unsupervised data-centric approach for exploring designs is to find clusters of similar designs and recommend only the cluster representatives for review. Design similarity can be defined not only on a purely functional level but also based on geometric properties, such as size, shape, and topology. While metrics such as chamfer distance measure the geometrical differences intuitively, it is more useful for design exploration to use metrics based on geometric features, which are extracted from high-dimensional 3D geometric data using dimensionality reduction techniques. If the Euclidean distance in the geometric features is meaningful, the features can be combined with performance attributes resulting in an aggregate feature vector that can potentially be useful in design exploration based on both geometry and performance. We propose a novel approach to evaluate such derived metrics by measuring their similarity with the metrics commonly used in 3D object classification. Furthermore, we measure clustering accuracy, which is a state-of-the-art unsupervised approach to evaluate metrics. For this purpose, we use a labeled, synthetic dataset with topologically complex designs. From our results, we conclude that Pointcloud Autoencoder is promising in encoding geometric features and developing a comprehensive design exploration method.
- Open Access
- 28 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Evolutionary Computation
- Vol. 26 (6) , 1221-1235
CarHoods10k: An Industry-Grade Data Set for Representation Learning and Design Optimization in Engineering Applications
Large-scale, high-quality data sets are central to the development of advanced machine learning techniques that increase the effectiveness of existing optimization methods or even inspire novel ones. Especially in the engineering domain such high quality data sets are rare due to confidentiality concerns and generation costs, be it computational or manual efforts. We therefore introduce the OSU-Honda Automobile Hood Dataset (CarHoods10k), an industry-grade 3D vehicle hood data set of over 10000 shapes along with mechanical performance data that were validated against real-world hood designs by industry experts. CarHoods10k offers researchers and practitioners the unique opportunity to develop novel methods on realistic data with relevance to real-world vehicle design. To illustrating central use cases, we first apply methods from geometric deep learning to learn a compact latent representation for design space exploration. Second, we use machine learning models to predict mechanical hood performance from the learned latent representation. We thus demonstrate the effectiveness of machine learning for building metamodels, which are used in design optimization whenever possible to replace costly engineering simulations. Third, we integrate CarHoods10k in a topology optimization approach based on evolutionary algorithms to demonstrate its capability to search for high-performing structures, while maintaining manufacturability constraints.
- 28 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Evolutionary Computation
- Vol. 26 (6) , 1221-1235
Open AccessCarHoods10k: An Industry-Grade Data Set for Representation Learning and Design Optimization in Engineering Applications
Large-scale, high-quality data sets are central to the development of advanced machine learning techniques that increase the effectiveness of existing optimization methods or even inspire novel ones. Especially in the engineering domain such high quality data sets are rare due to confidentiality concerns and generation costs, be it computational or manual efforts. We therefore introduce the OSU-Honda Automobile Hood Dataset (CarHoods10k), an industry-grade 3D vehicle hood data set of over 10000 shapes along with mechanical performance data that were validated against real-world hood designs by industry experts. CarHoods10k offers researchers and practitioners the unique opportunity to develop novel methods on realistic data with relevance to real-world vehicle design. To illustrating central use cases, we first apply methods from geometric deep learning to learn a compact latent representation for design space exploration. Second, we use machine learning models to predict mechanical hood performance from the learned latent representation. We thus demonstrate the effectiveness of machine learning for building metamodels, which are used in design optimization whenever possible to replace costly engineering simulations. Third, we integrate CarHoods10k in a topology optimization approach based on evolutionary algorithms to demonstrate its capability to search for high-performing structures, while maintaining manufacturability constraints.
- 18 October 2021
- Published by Springer Nature in Structural and Multidisciplinary Optimization
- Vol. 64 (6) , 4207
Correction to: Topology Optimization of 3D‑printed joints under crash loads using Evolutionary Algorithms
- 18 October 2021
- Published by Springer Nature in Structural and Multidisciplinary Optimization
- Vol. 64 (6) , 4207
Correction to: Topology Optimization of 3D‑printed joints under crash loads using Evolutionary Algorithms