Showing 1–7 of 7 results for author: Berenguer, A

The STOIC2021 COVID-19 AI challenge: applying reusable training methodologies to private data

Authors: Luuk H. Boulogne, Julian Lorenz, Daniel Kienzle, Robin Schon, Katja Ludwig, Rainer Lienhart, Simon Jegou, Guang Li, Cong Chen, Qi Wang, Derik Shi, Mayug Maniparambil, Dominik Muller, Silvan Mertes, Niklas Schroter, Fabio Hellmann, Miriam Elia, Ine Dirks, Matias Nicolas Bossa, Abel Diaz Berenguer, Tanmoy Mukherjee, Jef Vandemeulebroucke, Hichem Sahli, Nikos Deligiannis, Panagiotis Gonidakis , et al. (13 additional authors not shown)

Abstract: Challenges drive the state-of-the-art of automated medical image analysis. The quantity of public training data that they provide can limit the performance of their solutions. Public access to the training methodology for these solutions remains absent. This study implements the Type Three (T3) challenge format, which allows for training solutions on private data and guarantees reusable training m… ▽ More Challenges drive the state-of-the-art of automated medical image analysis. The quantity of public training data that they provide can limit the performance of their solutions. Public access to the training methodology for these solutions remains absent. This study implements the Type Three (T3) challenge format, which allows for training solutions on private data and guarantees reusable training methodologies. With T3, challenge organizers train a codebase provided by the participants on sequestered training data. T3 was implemented in the STOIC2021 challenge, with the goal of predicting from a computed tomography (CT) scan whether subjects had a severe COVID-19 infection, defined as intubation or death within one month. STOIC2021 consisted of a Qualification phase, where participants developed challenge solutions using 2000 publicly available CT scans, and a Final phase, where participants submitted their training methodologies with which solutions were trained on CT scans of 9724 subjects. The organizers successfully trained six of the eight Final phase submissions. The submitted codebases for training and running inference were released publicly. The winning solution obtained an area under the receiver operating characteristic curve for discerning between severe and non-severe COVID-19 of 0.815. The Final phase solutions of all finalists improved upon their Qualification phase solutions.HSUXJM-TNZF9CHSUXJM-TNZF9C △ Less

Submitted 25 June, 2023; v1 submitted 18 June, 2023; originally announced June 2023.

Representation Learning with Information Theory for COVID-19 Detection

Authors: Abel Díaz Berenguer, Tanmoy Mukherjee, Matias Bossa, Nikos Deligiannis, Hichem Sahli

Abstract: Successful data representation is a fundamental factor in machine learning based medical imaging analysis. Deep Learning (DL) has taken an essential role in robust representation learning. However, the inability of deep models to generalize to unseen data can quickly overfit intricate patterns. Thereby, we can conveniently implement strategies to aid deep models in discovering useful priors from d… ▽ More Successful data representation is a fundamental factor in machine learning based medical imaging analysis. Deep Learning (DL) has taken an essential role in robust representation learning. However, the inability of deep models to generalize to unseen data can quickly overfit intricate patterns. Thereby, we can conveniently implement strategies to aid deep models in discovering useful priors from data to learn their intrinsic properties. Our model, which we call a dual role network (DRN), uses a dependency maximization approach based on Least Squared Mutual Information (LSMI). The LSMI leverages dependency measures to ensure representation invariance and local smoothness. While prior works have used information theory measures like mutual information, known to be computationally expensive due to a density estimation step, our LSMI formulation alleviates the issues of intractable mutual information estimation and can be used to approximate it. Experiments on CT based COVID-19 Detection and COVID-19 Severity Detection benchmarks demonstrate the effectiveness of our method. △ Less

Submitted 4 July, 2022; originally announced July 2022.

Explainable-by-design Semi-Supervised Representation Learning for COVID-19 Diagnosis from CT Imaging

Authors: Abel Díaz Berenguer, Hichem Sahli, Boris Joukovsky, Maryna Kvasnytsia, Ine Dirks, Mitchel Alioscha-Perez, Nikos Deligiannis, Panagiotis Gonidakis, Sebastián Amador Sánchez, Redona Brahimetaj, Evgenia Papavasileiou, Jonathan Cheung-Wai Chana, Fei Li, Shangzhen Song, Yixin Yang, Sofie Tilborghs, Siri Willems, Tom Eelbode, Jeroen Bertels, Dirk Vandermeulen, Frederik Maes, Paul Suetens, Lucas Fidon, Tom Vercauteren, David Robben , et al. (15 additional authors not shown)

Abstract: Our motivating application is a real-world problem: COVID-19 classification from CT imaging, for which we present an explainable Deep Learning approach based on a semi-supervised classification pipeline that employs variational autoencoders to extract efficient feature embedding. We have optimized the architecture of two different networks for CT images: (i) a novel conditional variational autoenc… ▽ More Our motivating application is a real-world problem: COVID-19 classification from CT imaging, for which we present an explainable Deep Learning approach based on a semi-supervised classification pipeline that employs variational autoencoders to extract efficient feature embedding. We have optimized the architecture of two different networks for CT images: (i) a novel conditional variational autoencoder (CVAE) with a specific architecture that integrates the class labels inside the encoder layers and uses side information with shared attention layers for the encoder, which make the most of the contextual clues for representation learning, and (ii) a downstream convolutional neural network for supervised classification using the encoder structure of the CVAE. With the explainable classification results, the proposed diagnosis system is very effective for COVID-19 classification. Based on the promising results obtained qualitatively and quantitatively, we envisage a wide deployment of our developed technique in large-scale clinical studies.Code is available at https://git.etrovub.be/AVSP/ct-based-covid-19-diagnostic-tool.git. △ Less

Submitted 2 September, 2021; v1 submitted 23 November, 2020; originally announced November 2020.

Comparative study of multipactor effect in rectangular and parallel-plate waveguides partially loaded with dielectric

Authors: A. Berenguer, A. Coves, E. Bronchalo

Abstract: In this work, a comparative study of the susceptibility chart in a partially dielectric-loaded rectangular waveguide and in its equivalent partially dielectric-loaded parallel-plate waveguide with the same height is performed. This study shows how the inhomogeneity of the electric field inside the rectangular waveguide modifies the edges of the multipactor susceptibility chart with respect to that… ▽ More In this work, a comparative study of the susceptibility chart in a partially dielectric-loaded rectangular waveguide and in its equivalent partially dielectric-loaded parallel-plate waveguide with the same height is performed. This study shows how the inhomogeneity of the electric field inside the rectangular waveguide modifies the edges of the multipactor susceptibility chart with respect to that predicted by the parallel-plate waveguide case. Moreover, comparisons of the evolution of the population and the trajectory of the effective electron and the DC field appearing in the waveguide in a point well inside the multipactor region obtained with both models are performed, showing some quantitative differences in the multipactor evolution, but a similar qualitative behaviour is observed. △ Less

Submitted 2 May, 2020; originally announced May 2020.

Analysis of multipactor effect in parallel-plate and rectangular waveguides

Authors: A. Berenguer, A. Coves, E. Bronchalo, B. Gimeno, V. Boria

Abstract: In this work it is investigated, in the parallel-plate waveguide case, how the 1D, 2D or 3D motion of the electrons inside the waveguide can affect the generalized susceptibility diagrams, by means of a developed model capable of tracking the exact trajectory of multiple effective electrons which includes effects such as the spreading of the secondary electron departure kinetic energies or the dep… ▽ More In this work it is investigated, in the parallel-plate waveguide case, how the 1D, 2D or 3D motion of the electrons inside the waveguide can affect the generalized susceptibility diagrams, by means of a developed model capable of tracking the exact trajectory of multiple effective electrons which includes effects such as the spreading of the secondary electron departure kinetic energies or the dependence on elastic and inelastic electrons. On the other hand, a comparative study of the susceptibility charts in a parallel-plate and in its equivalent rectangular waveguide with the same height is performed, showing how the inhomogeneity of the electric field inside the waveguide modifies the multipactor region with respect to that predicted by the parallel-plate waveguide case. The results of this study are going to be extended to a partially dielectric-loaded rectangular waveguide, which is a problem of great interest in the space industry that has not yet been rigorously investigated in the literature. △ Less

Submitted 2 May, 2020; originally announced May 2020.

Analysis of mathematical techniques for the calculation of the electrostatic field in a dielectric-loaded waveguide

Authors: A. Berenguer, A. Coves, E. Bronchalo, F. Mesa, B. Gimeno

Abstract: The resolution of the Green's function for obtaining the electrostatic potential generated by the charges located in the dielectric layer of a rectangular waveguide requires efficient integration techniques. Due to the characteristics and the oscillatory nature of the function to be integrated, the use of the Filon's method together with a convergence analysis is adequate. However, the application… ▽ More The resolution of the Green's function for obtaining the electrostatic potential generated by the charges located in the dielectric layer of a rectangular waveguide requires efficient integration techniques. Due to the characteristics and the oscillatory nature of the function to be integrated, the use of the Filon's method together with a convergence analysis is adequate. However, the application of this numerical integration technique can lead to numerical instabilities in the result. For this reason, in this research work we have presented two different methods to deal with these numerical errors of integration: SSA/MSSA and GPR. In this way it is possible to clean the electrostatic potential avoiding subsequent errors in the calculation of the electrostatic field. △ Less

Submitted 2 May, 2020; originally announced May 2020.

Comments: arXiv admin note: text overlap with arXiv:2005.00909

Analysis of the electrostratic field generated by a charge distribution on a dielectric layer loading a rectangular waveguide

Authors: A. Berenguer, A. Coves, F. Mesa, E. Bronchalo, B. Gimeno, V. Boria

Abstract: The goal of this paper is to study the electrostatic field due to an arbitrary charge distribution on a dielectric layer in a dielectric-loaded rectangular waveguide. In order to obtain this electrostatic field, the potential due to a point charge on the dielectric layer is solved in advance. The high computational complexity of this problem requires the use of different numerical integration tech… ▽ More The goal of this paper is to study the electrostatic field due to an arbitrary charge distribution on a dielectric layer in a dielectric-loaded rectangular waveguide. In order to obtain this electrostatic field, the potential due to a point charge on the dielectric layer is solved in advance. The high computational complexity of this problem requires the use of different numerical integration techniques (e.g. Filon, Gauss-Kronrod, Lobatto, ...) and interpolation methods. Using the principle of superposition, the potential due to an arbitrary charge distribution on a dielectric layer is obtained by adding the individual contribution of each point charge. Finally, a numerical differentiation of the potential is carried out to obtain the electrostatic field in the waveguide. The results of this electrostatic problem are going to be extended to model the multipactor effect, which is a problem of great interest in the space industry. △ Less

Submitted 2 May, 2020; originally announced May 2020.