1 code implementation • NeurIPS 2023 • Kyriakos Flouris, Ender Konukoglu
Alternatively, if a locally orthogonal and/or sparse basis is to be learned, here coined canonical intrinsic basis, it can serve in learning a more compact latent space representation.
no code implementations • 3 Aug 2023 • Katarína Tóthová, Ľubor Ladický, Daniel Thul, Marc Pollefeys, Ender Konukoglu
Predictive variability due to data ambiguities has typically been addressed via construction of dedicated models with built-in probabilistic capabilities that are trained to predict uncertainty estimates as variables of interest.
no code implementations • 12 Apr 2023 • Gustav Bredell, Kyriakos Flouris, Krishna Chaitanya, Ertunc Erdil, Ender Konukoglu
Variational autoencoders (VAEs) are powerful generative modelling methods, however they suffer from blurry generated samples and reconstructions compared to the images they have been trained on.
no code implementations • 31 Mar 2023 • Gary Sarwin, Alessandro Carretta, Victor Staartjes, Matteo Zoli, Diego Mazzatenta, Luca Regli, Carlo Serra, Ender Konukoglu
With this motivation, we present a method for live image-only guidance leveraging a large data set of annotated neurosurgical videos. First, we report the performance of a deep learning-based object detection method, YOLO, on detecting anatomical structures in neurosurgical images.
no code implementations • 20 Feb 2023 • Jiahua Dong, Yang Cong, Gan Sun, Lixu Wang, Lingjuan Lyu, Jun Li, Ender Konukoglu
Moreover, they cannot explore which 3D geometric characteristics are essential to alleviate the catastrophic forgetting on old classes of 3D objects.
1 code implementation • 30 Jan 2023 • Tianfei Zhou, Ender Konukoglu
To reach this goal, we propose FedFA to tackle federated learning from a distinct perspective of federated feature augmentation.
no code implementations • 28 Nov 2022 • Jakob Geusen, Gustav Bredell, Tianfei Zhou, Ender Konukoglu
Partitioning an image into superpixels based on the similarity of pixels with respect to features such as colour or spatial location can significantly reduce data complexity and improve subsequent image processing tasks.
1 code implementation • 13 Apr 2022 • Edoardo Mello Rella, Ajad Chhatkuli, Ender Konukoglu, Luc van Gool
With neural networks, several other variations and training principles have been proposed with the goal to represent all classes of shapes.
1 code implementation • CVPR 2022 • Tianfei Zhou, Wenguan Wang, Ender Konukoglu, Luc van Gool
Prevalent semantic segmentation solutions, despite their different network designs (FCN based or attention based) and mask decoding strategies (parametric softmax based or pixel-query based), can be placed in one category, by considering the softmax weights or query vectors as learnable class prototypes.
1 code implementation • ICLR 2022 • Edoardo Mello Rella, Ajad Chhatkuli, Yun Liu, Ender Konukoglu, Luc van Gool
One of the key problems in boundary detection is the label representation, which typically leads to class imbalance and, as a consequence, to thick boundaries that require non-differential post-processing steps to be thinned.
1 code implementation • 10 Feb 2022 • Neerav Karani, Georg Brunner, Ertunc Erdil, Simin Fei, Kerem Tezcan, Krishna Chaitanya, Ender Konukoglu
We use 1D marginal distributions of a trained task CNN's features as experts in the FoE model.
1 code implementation • 19 Dec 2021 • Gustav Bredell, Ertunc Erdil, Bruno Weber, Ender Konukoglu
In addition, the image generator reproduces low-frequency features of the deconvolved image faster than that of a blurry image.
no code implementations • 17 Dec 2021 • Krishna Chaitanya, Ertunc Erdil, Neerav Karani, Ender Konukoglu
In this paper, we propose a local contrastive loss to learn good pixel level features useful for segmentation by exploiting semantic label information obtained from pseudo-labels of unlabeled images alongside limited annotated images.
1 code implementation • CVPR 2022 • Metin Ersin Arican, Ozgur Kara, Gustav Bredell, Ender Konukoglu
Our experiments show that image-specific metrics can reduce the search space to a small cohort of models, of which the best model outperforms current NAS approaches for image restoration.
1 code implementation • 20 Jun 2021 • Tianfei Zhou, Liulei Li, Gustav Bredell, Jianwu Li, Ender Konukoglu
The proposed network has two appealing characteristics: 1) The memory-augmented network offers the ability to quickly encode past segmentation information, which will be retrieved for the segmentation of other slices; 2) The quality assessment module enables the model to directly estimate the qualities of segmentation predictions, which allows an active learning paradigm where users preferentially label the lowest-quality slice for multi-round refinement.
no code implementations • 11 May 2021 • Kyriakos Flouris, Anna Volokitin, Gustav Bredell, Ender Konukoglu
In this work, we investigate a decoder-only method that uses gradient flow to encode data samples in the latent space.
no code implementations • 30 Mar 2021 • Alexis Perakis, Ali Gorji, Samriddhi Jain, Krishna Chaitanya, Simone Rizza, Ender Konukoglu
Learning robust representations to discriminate cell phenotypes based on microscopy images is important for drug discovery.
1 code implementation • NeurIPS 2021 • Sara Sangalli, Ertunc Erdil, Andreas Hoetker, Olivio Donati, Ender Konukoglu
Such class imbalance is ubiquitous in clinical applications and very crucial to handle because the classes with fewer samples most often correspond to critical cases (e. g., cancer) where misclassifications can have severe consequences.
5 code implementations • ICCV 2021 • Wenguan Wang, Tianfei Zhou, Fisher Yu, Jifeng Dai, Ender Konukoglu, Luc van Gool
Inspired by the recent advance in unsupervised contrastive representation learning, we propose a pixel-wise contrastive framework for semantic segmentation in the fully supervised setting.
2 code implementations • 19 Jan 2021 • Ke Li, Dengxin Dai, Ender Konukoglu, Luc van Gool
With these contributions, our method is able to learn from heterogeneous datasets and lift the requirement for having a large amount of HD HSI training samples.
no code implementations • 5 Oct 2020 • Katarína Tóthová, Sarah Parisot, Matthew Lee, Esther Puyol-Antón, Andrew King, Marc Pollefeys, Ender Konukoglu
Surface reconstruction from magnetic resonance (MR) imaging data is indispensable in medical image analysis and clinical research.
1 code implementation • 30 Sep 2020 • Kerem C. Tezcan, Neerav Karani, Christian F. Baumgartner, Ender Konukoglu
In this paper, we propose a method that instead returns multiple images which are possible under the acquisition model and the chosen prior to capture the uncertainty in the inversion process.
1 code implementation • 16 Aug 2020 • Marc Gantenbein, Ertunc Erdil, Ender Konukoglu
We incorporate the reversible blocks into a recently proposed architecture called PHiSeg that is developed for uncertainty quantification in medical image segmentation.
no code implementations • 26 Jul 2020 • Mélanie Gaillochet, Kerem C. Tezcan, Ender Konukoglu
To this end, we use an unsupervised learning based reconstruction algorithm as our basis and combine it with a N4-based bias field estimation method, in a joint optimization scheme.
1 code implementation • 9 Jul 2020 • Anna Volokitin, Ertunc Erdil, Neerav Karani, Kerem Can Tezcan, Xiaoran Chen, Luc van Gool, Ender Konukoglu
We propose a method to model 3D MR brain volumes distribution by combining a 2D slice VAE with a Gaussian model that captures the relationships between slices.
1 code implementation • 9 Jul 2020 • Krishna Chaitanya, Neerav Karani, Christian F. Baumgartner, Ertunc Erdil, Anton Becker, Olivio Donati, Ender Konukoglu
In this work, we propose a novel task-driven data augmentation method for learning with limited labeled data where the synthetic data generator, is optimized for the segmentation task.
1 code implementation • NeurIPS 2020 • Krishna Chaitanya, Ertunc Erdil, Neerav Karani, Ender Konukoglu
In this work, we propose strategies for extending the contrastive learning framework for segmentation of volumetric medical images in the semi-supervised setting with limited annotations, by leveraging domain-specific and problem-specific cues.
1 code implementation • 18 Jun 2020 • Ertunc Erdil, Krishna Chaitanya, Neerav Karani, Ender Konukoglu
The results demonstrate that the proposed method consistently achieves high OOD detection performance in both classification and segmentation tasks and improves state-of-the-art in almost all cases.
no code implementations • 30 Apr 2020 • Xiaoran Chen, Suhang You, Kerem Can Tezcan, Ender Konukoglu
In this work, we approach unsupervised lesion detection as an image restoration problem and propose a probabilistic model that uses a network-based prior as the normative distribution and detect lesions pixel-wise using MAP estimation.
2 code implementations • 9 Apr 2020 • Neerav Karani, Ertunc Erdil, Krishna Chaitanya, Ender Konukoglu
In medical image segmentation, this premise is violated when there is a mismatch between training and test images in terms of their acquisition details, such as the scanner model or the protocol.
no code implementations • 31 Jan 2020 • Esther Puyol Anton, Bram Ruijsink, Christian F. Baumgartner, Matthew Sinclair, Ender Konukoglu, Reza Razavi, Andrew P. King
The PHiSeg network and QC were validated against manual analysis on a cohort of the UK Biobank containing healthy subjects and chronic cardiomyopathy patients.
no code implementations • 10 Oct 2019 • Ben Glocker, Robert Robinson, Daniel C. Castro, Qi Dou, Ender Konukoglu
This is an empirical study to investigate the impact of scanner effects when using machine learning on multi-site neuroimaging data.
4 code implementations • 14 Jun 2019 • Robin Brügger, Christian F. Baumgartner, Ender Konukoglu
Increasing network depth led to higher segmentation accuracy while growing the memory footprint only by a very small fraction, thanks to the partially reversible architecture.
3 code implementations • 7 Jun 2019 • Christian F. Baumgartner, Kerem C. Tezcan, Krishna Chaitanya, Andreas M. Hötker, Urs J. Muehlematter, Khoschy Schawkat, Anton S. Becker, Olivio Donati, Ender Konukoglu
Segmentation of anatomical structures and pathologies is inherently ambiguous.
no code implementations • 21 May 2019 • M. Jorge Cardoso, Aasa Feragen, Ben Glocker, Ender Konukoglu, Ipek Oguz, Gozde Unal, Tom Vercauteren
This compendium gathers all the accepted extended abstracts from the Second International Conference on Medical Imaging with Deep Learning (MIDL 2019), held in London, UK, 8-10 July 2019.
1 code implementation • 20 Feb 2019 • Lukas Jendele, Ondrej Skopek, Anton S. Becker, Ender Konukoglu
Supervised deep learning relies on the assumption that enough training data is available, which presents a problem for its application to several fields, like medical imaging.
1 code implementation • 11 Feb 2019 • Krishna Chaitanya, Neerav Karani, Christian Baumgartner, Olivio Donati, Anton Becker, Ender Konukoglu
However, there is potential to improve the approach by (i) explicitly modeling deformation fields (non-affine spatial transformation) and intensity transformations and (ii) leveraging unlabelled data during the generative process.
2 code implementations • 19 Nov 2018 • Anton S. Becker, Lukas Jendele, Ondrej Skopek, Nicole Berger, Soleen Ghafoor, Magda Marcon, Ender Konukoglu
At the higher resolution, all radiologists showed significantly lower detection rate of cancer in the modified images (0. 77-0. 84 vs. 0. 59-0. 69, p=0. 008), however, they were now able to reliably detect modified images due to better visibility of artifacts (0. 92, 0. 92 and 0. 97).
1 code implementation • ICLR 2019 • Daniel C. Castro, Jeremy Tan, Bernhard Kainz, Ender Konukoglu, Ben Glocker
Revealing latent structure in data is an active field of research, having introduced exciting technologies such as variational autoencoders and adversarial networks, and is essential to push machine learning towards unsupervised knowledge discovery.
no code implementations • 30 Jul 2018 • Katarína Tóthová, Sarah Parisot, Matthew C. H. Lee, Esther Puyol-Antón, Lisa M. Koch, Andrew P. King, Ender Konukoglu, Marc Pollefeys
Surface reconstruction is a vital tool in a wide range of areas of medical image analysis and clinical research.
no code implementations • 24 Jul 2018 • Jana Kemnitz, Christian F. Baumgartner, Wolfgang Wirth, Felix Eckstein, Sebastian K. Eder, Ender Konukoglu
In this work we propose a cost function which allows integration of multiple datasets with heterogeneous label subsets into a joint training.
no code implementations • 23 Jul 2018 • Gustav Bredell, Christine Tanner, Ender Konukoglu
Automatic segmentation has great potential to facilitate morphological measurements while simultaneously increasing efficiency.
no code implementations • 19 Jul 2018 • Christine Tanner, Firat Ozdemir, Romy Profanter, Valeriy Vishnevsky, Ender Konukoglu, Orcun Goksel
Performance for the abdominal region was similar to that of CT-MRI NMI registration (77. 4 vs. 78. 8%) when using 3D synthesizing MRIs (12 slices) and medium sized receptive fields for the discriminator.
no code implementations • 12 Jul 2018 • Yigit B. Can, Krishna Chaitanya, Basil Mustafa, Lisa M. Koch, Ender Konukoglu, Christian F. Baumgartner
We find that the networks trained on scribbles suffer from a remarkably small degradation in Dice of only 2. 9% (cardiac) and 4. 5% (prostate) with respect to a network trained on full annotations.
no code implementations • 14 Jun 2018 • Xiaoran Chen, Nick Pawlowski, Martin Rajchl, Ben Glocker, Ender Konukoglu
In this paper, we explore the feasibility of using state-of-the-art auto-encoder-based deep generative models, such as variational and adversarial auto-encoders, for one such task: abnormality detection in medical imaging.
1 code implementation • 13 Jun 2018 • Xiaoran Chen, Ender Konukoglu
Lesion detection in brain Magnetic Resonance Images (MRI) remains a challenging task.
1 code implementation • 25 May 2018 • Neerav Karani, Krishna Chaitanya, Christian Baumgartner, Ender Konukoglu
We evaluate the method for brain structure segmentation in MR images.
1 code implementation • 12 Apr 2018 • Lin Zhang, Neerav Karani, Christine Tanner, Ender Konukoglu
Temporal interpolation of navigator slices an be used to reduce the number of navigator acquisitions without degrading specificity in stacking.
no code implementations • 30 Nov 2017 • Kerem C. Tezcan, Christian F. Baumgartner, Roger Luechinger, Klaas P. Pruessmann, Ender Konukoglu
Deep learning (DL) provides a powerful framework for extracting such information from existing image datasets, through learning, and then using it for reconstruction.
3 code implementations • CVPR 2018 • Christian F. Baumgartner, Lisa M. Koch, Kerem Can Tezcan, Jia Xi Ang, Ender Konukoglu
Attributing the pixels of an input image to a certain category is an important and well-studied problem in computer vision, with applications ranging from weakly supervised localisation to understanding hidden effects in the data.
1 code implementation • 13 Sep 2017 • Christian F. Baumgartner, Lisa M. Koch, Marc Pollefeys, Ender Konukoglu
Accurate segmentation of the heart is an important step towards evaluating cardiac function.
no code implementations • 29 Aug 2017 • Mehmet Turan, Yusuf Yigit Pilavci, Ipek Ganiyusufoglu, Helder Araujo, Ender Konukoglu, Metin Sitti
Since the development of capsule endoscopcy technology, substantial progress were made in converting passive capsule endoscopes to robotic active capsule endoscopes which can be controlled by the doctor.
no code implementations • 22 Aug 2017 • Mehmet Turan, Yasin Almalioglu, Helder Araujo, Ender Konukoglu, Metin Sitti
Ingestible wireless capsule endoscopy is an emerging minimally invasive diagnostic technology for inspection of the GI tract and diagnosis of a wide range of diseases and pathologies.
no code implementations • 18 May 2017 • Mehmet Turan, Yusuf Yigit Pilavci, Redhwan Jamiruddin, Helder Araujo, Ender Konukoglu, Metin Sitti
In the gastrointestinal (GI) tract endoscopy field, ingestible wireless capsule endoscopy is emerging as a novel, minimally invasive diagnostic technology for inspection of the GI tract and diagnosis of a wide range of diseases and pathologies.
no code implementations • 17 May 2017 • Mehmet Turan, Yasin Almalioglu, Hunter Gilbert, Helder Araujo, Ender Konukoglu, Metin Sitti
A reliable, real-time simultaneous localization and mapping (SLAM) method is crucial for the navigation of actively controlled capsule endoscopy robots.
no code implementations • 15 May 2017 • Mehmet Turan, Yasin Almalioglu, Ender Konukoglu, Metin Sitti
We present a robust deep learning based 6 degrees-of-freedom (DoF) localization system for endoscopic capsule robots.
no code implementations • 15 May 2017 • Mehmet Turan, Yasin Almalioglu, Helder Araujo, Ender Konukoglu, Metin Sitti
In this paper, we propose to our knowledge for the first time in literature a visual simultaneous localization and mapping (SLAM) method specifically developed for endoscopic capsule robots.
1 code implementation • 10 Jan 2017 • Ender Konukoglu, Ben Glocker
Analyses on the ADNI dataset show that RSM can also improve correlation between subject-specific detections in cortical thickness data and non-imaging markers of Alzheimer's Disease (AD), such as the Mini Mental State Examination Score and Cerebrospinal Fluid amyloid-$\beta$ levels.
1 code implementation • 10 Oct 2014 • Ender Konukoglu, Melanie Ganz
Random Forest has become one of the most popular tools for feature selection.