Georgios K. Papageorgiou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Georgios K. Papageorgiou has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
Constantinos B. Papadias
The American College of Greece,Athens,GR
9 shared publicationsKonstantinos Ntougias
Broadband Wireless and Sensor Networks Laboratory, Athens Information Technology, Marousi, Greece
4 shared publicationsPantelis Bouboulis
Department of Informatics and Telecommunications, University of Athens Athens, Greece, 157 84
4 shared publicationsSergios Theodoridis
Department of Informatics and Telecommunications, University of Athens Athens, Greece, 157 84
4 shared publicationsDimitrios Ntaikos
Athens Information Technology, 44 Kifissias Ave. Athens, 15125, Greece
3 shared publications- Lachlan Arthur
School of Engineering & Physical Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, Scotland, United Kingdom
3 shared publications
Latest Publications
- Open Access
- 6 December 2024
- Published by Elsevier in Journal of Theoretical Biology
- Vol. 599, 112018
Super-resolution ultrasound imaging of ischaemia flow: An in silico study
- 6 December 2024
- Published by Elsevier in Journal of Theoretical Biology
- Vol. 599, 112018
Open AccessSuper-resolution ultrasound imaging of ischaemia flow: An in silico study
- Open Access
- 22 November 2024
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 69 (23) , 235006
Comparison of contrast-enhanced ultrasound imaging (CEUS) and super-resolution ultrasound (SRU) for the quantification of ischaemia flow redistribution: a theoretical study
The study of microcirculation can reveal important information related to pathology. Focusing on alterations that are represented by an obstruction of blood flow in microcirculatory regions may provide an insight into vascular biomarkers. The current in silico study assesses the capability of contrast enhanced ultrasound (CEUS) and super-resolution ultrasound imaging (SRU) flow-quantification to study occlusive actions in a microvascular bed, particularly the ability to characterise known and model induced flow behaviours. The aim is to investigate theoretical limits with the use of CEUS and SRU in order to propose realistic biomarker targets relevant for clinical diagnosis. Results from CEUS flow parameters display limitations congruent with prior investigations. Conventional resolution limits lead to signals dominated by large vessels, making discrimination of microvasculature specific signals difficult. Additionally, some occlusions lead to weakened parametric correlation against flow rate in the remainder of the network. Loss of correlation is dependent on the degree to which flow is redistributed, with comparatively minor redistribution correlating in accordance with ground truth measurements for change in mean transit time, dMTT (CEUS, R = 0.85; GT, R = 0.82) and change in peak intensity, d I p (CEUS, R = 0.87; GT, R = 0.96). Major redistributions, however, result in a loss of correlation, demonstrating that the effectiveness of time-intensity curve parameters is influenced by the site of occlusion. Conversely, results from SRU processing provides accurate depiction of the anatomy and dynamics present in the vascular bed, that extends to individual microvessels. Correspondence between model vessel structure displayed in SRU maps with the ground truth was > 91 % for cases of minor and major flow redistributions. In conclusion, SRU appears to be a highly promising technology in the quantification of subtle flow phenomena due ischaemia induced vascular flow redistribution.
- 22 November 2024
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 69 (23) , 235006
Open AccessComparison of contrast-enhanced ultrasound imaging (CEUS) and super-resolution ultrasound (SRU) for the quantification of ischaemia flow redistribution: a theoretical study
The study of microcirculation can reveal important information related to pathology. Focusing on alterations that are represented by an obstruction of blood flow in microcirculatory regions may provide an insight into vascular biomarkers. The current in silico study assesses the capability of contrast enhanced ultrasound (CEUS) and super-resolution ultrasound imaging (SRU) flow-quantification to study occlusive actions in a microvascular bed, particularly the ability to characterise known and model induced flow behaviours. The aim is to investigate theoretical limits with the use of CEUS and SRU in order to propose realistic biomarker targets relevant for clinical diagnosis. Results from CEUS flow parameters display limitations congruent with prior investigations. Conventional resolution limits lead to signals dominated by large vessels, making discrimination of microvasculature specific signals difficult. Additionally, some occlusions lead to weakened parametric correlation against flow rate in the remainder of the network. Loss of correlation is dependent on the degree to which flow is redistributed, with comparatively minor redistribution correlating in accordance with ground truth measurements for change in mean transit time, dMTT (CEUS, R = 0.85; GT, R = 0.82) and change in peak intensity, d I p (CEUS, R = 0.87; GT, R = 0.96). Major redistributions, however, result in a loss of correlation, demonstrating that the effectiveness of time-intensity curve parameters is influenced by the site of occlusion. Conversely, results from SRU processing provides accurate depiction of the anatomy and dynamics present in the vascular bed, that extends to individual microvessels. Correspondence between model vessel structure displayed in SRU maps with the ground truth was > 91 % for cases of minor and major flow redistributions. In conclusion, SRU appears to be a highly promising technology in the quantification of subtle flow phenomena due ischaemia induced vascular flow redistribution.
- Open Access
- 30 August 2024
- Published by Elsevier in Ultrasonics
- Vol. 145, 107451
Vessel recovery using ultrasound localisation microscopy: An in silico comparative study between minimum variance and delay-and-sum beamformers
- 30 August 2024
- Published by Elsevier in Ultrasonics
- Vol. 145, 107451
Open AccessVessel recovery using ultrasound localisation microscopy: An in silico comparative study between minimum variance and delay-and-sum beamformers
- Top Cited PapersOpen Access
- 16 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 69 (1053587X) , 3714-3729
Deep Networks for Direction-of-Arrival Estimation in Low SNR
In this work, we consider direction-of-arrival (DoA) estimation in the presence of extreme noise using Deep Learning (DL). In particular, we introduce a Convolutional Neural Network (CNN) that predicts angular directions using the sample covariance matrix estimate. The network is trained from multi-channel data of the true array manifold matrix in the low signal-to-noise-ratio (SNR) regime. By adopting an on-grid approach, we model the problem as a multi-label classification task and train the CNN to predict DoAs across all SNRs. The proposed architecture demonstrates enhanced robustness in the presence of noise, and resilience to a relatively small number of snapshots. Moreover, it is able to resolve angles within the grid resolution. Experimental results demonstrate significant performance gains in the low-SNR regime compared to state-of-the-art methods and without the requirement of any parameter tuning in both cases of correlated and uncorrelated sources. Finally, we relax the assumption that the number of sources is known a priori and present a training method, where the CNN learns to infer their number and predict the DoAs with high confidence. The increased robustness of the proposed solution is highly desirable in challenging scenarios that arise in several fields, ranging from wireless array sensors to acoustic microphones or sonars.
- 16 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 69 (1053587X) , 3714-3729
Top Cited PapersOpen AccessDeep Networks for Direction-of-Arrival Estimation in Low SNR
In this work, we consider direction-of-arrival (DoA) estimation in the presence of extreme noise using Deep Learning (DL). In particular, we introduce a Convolutional Neural Network (CNN) that predicts angular directions using the sample covariance matrix estimate. The network is trained from multi-channel data of the true array manifold matrix in the low signal-to-noise-ratio (SNR) regime. By adopting an on-grid approach, we model the problem as a multi-label classification task and train the CNN to predict DoAs across all SNRs. The proposed architecture demonstrates enhanced robustness in the presence of noise, and resilience to a relatively small number of snapshots. Moreover, it is able to resolve angles within the grid resolution. Experimental results demonstrate significant performance gains in the low-SNR regime compared to state-of-the-art methods and without the requirement of any parameter tuning in both cases of correlated and uncorrelated sources. Finally, we relax the assumption that the number of sources is known a priori and present a training method, where the CNN learns to infer their number and predict the DoAs with high confidence. The increased robustness of the proposed solution is highly desirable in challenging scenarios that arise in several fields, ranging from wireless array sensors to acoustic microphones or sonars.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
3D Beamforming with Multi-Active Multi-Passive Antenna Arrays Using Stochastic Optimization
In this paper, we consider the 3D beamforming with Multi-Active Multi-Passive (MAMP) antenna arrays. For the optimization of the hybrid array’s active elements’ (AEs) weights and passive elements’ (PEs) loads, we propose a novel algorithm, i.e., the alternating optimization - stochastic 3D beamforming algorithm (AO-S3DBA). The scheme is built on a generalized cost function and alternates between the optimization of the loads and weights. Despite the dramatic increase in data points, we managed to decrease the algorithm’s complexity by employing Adam, a popular accelerator for stochastic optimization, for the update of the optimization variables. We present simulation results, where the proposed MAMP array can successfully emulate the beam of a uniform rectangular array (URA) with the use of 50% less AEs and with beam steering capabilities towards various azimuth and elevation directions. We believe that this newly proposed type of hybrid transceiver offers a good trade-off between cost and performance, which can be particularly useful in many industrial and Internet-of-Things (IoT) applications, where the deployment of a large number of transceivers / sensors directly affects the application’s total cost.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
3D Beamforming with Multi-Active Multi-Passive Antenna Arrays Using Stochastic Optimization
In this paper, we consider the 3D beamforming with Multi-Active Multi-Passive (MAMP) antenna arrays. For the optimization of the hybrid array’s active elements’ (AEs) weights and passive elements’ (PEs) loads, we propose a novel algorithm, i.e., the alternating optimization - stochastic 3D beamforming algorithm (AO-S3DBA). The scheme is built on a generalized cost function and alternates between the optimization of the loads and weights. Despite the dramatic increase in data points, we managed to decrease the algorithm’s complexity by employing Adam, a popular accelerator for stochastic optimization, for the update of the optimization variables. We present simulation results, where the proposed MAMP array can successfully emulate the beam of a uniform rectangular array (URA) with the use of 50% less AEs and with beam steering capabilities towards various azimuth and elevation directions. We believe that this newly proposed type of hybrid transceiver offers a good trade-off between cost and performance, which can be particularly useful in many industrial and Internet-of-Things (IoT) applications, where the deployment of a large number of transceivers / sensors directly affects the application’s total cost.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Direction-of-Arrival Estimation in the Low-SNR Regime via a Denoising Autoencoder
The performance of covariance-based DoA estimation methods is limited in practice, particularly in the low signal-to-noise ratio (SNR) regime, due to the finite number of observations. In this work, we approach the direction-of-arrival (DoA) estimation in the presence of extreme noise from the Machine Learning (ML) perspective using Deep Learning (DL). First, we derive a relation between the covariance matrix and its sample estimate formulating the problem as a manifold learning task. Next, we train a denoising autoencoder (DAE) that predicts a Hermitian matrix, which is subsequently used for the DoA estimation. Experimental results demonstrate significant performance gains in terms of the root-mean-squared error (RMSE) in the low-SNR regime by using popular covariance-based DoA estimators. Nevertheless, the proposed method runs independent of the DoA estimator, opening up new possibilities for the testing of other methods as well. We believe that the proposed approach has several applications, ranging from wireless array sensors to microphones and transducers used in ultrasound imaging, where the operating environments are characterized by extreme noise.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Direction-of-Arrival Estimation in the Low-SNR Regime via a Denoising Autoencoder
The performance of covariance-based DoA estimation methods is limited in practice, particularly in the low signal-to-noise ratio (SNR) regime, due to the finite number of observations. In this work, we approach the direction-of-arrival (DoA) estimation in the presence of extreme noise from the Machine Learning (ML) perspective using Deep Learning (DL). First, we derive a relation between the covariance matrix and its sample estimate formulating the problem as a manifold learning task. Next, we train a denoising autoencoder (DAE) that predicts a Hermitian matrix, which is subsequently used for the DoA estimation. Experimental results demonstrate significant performance gains in terms of the root-mean-squared error (RMSE) in the low-SNR regime by using popular covariance-based DoA estimators. Nevertheless, the proposed method runs independent of the DoA estimator, opening up new possibilities for the testing of other methods as well. We believe that the proposed approach has several applications, ranging from wireless array sensors to microphones and transducers used in ultrasound imaging, where the operating environments are characterized by extreme noise.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 4632-4636
Fast Direction-of-arrival Estimation of Multiple Targets Using Deep Learning and Sparse Arrays
In this work, we focus on improving the Direction-of-Arrival (DoA) estimation of multiple targets/sources from a small number of snapshots. Estimation via the sample covariance matrix is known to perform poorly, since the true manifold structure is not revealed for a small number of samples. First, we explicitly model the sample covariance matrix that is used for the DoA estimation as a noisy version of the true one. Next, we employ a stacked denoising autoencoder (DAE) that predicts a statistically "richer" version of the sampled matrix that is subsequently used for the DoA estimation. Moreover, we consider a limited number of sensors (comparable to the number of sources) in a non-uniform linear configuration and introduce an end-to-end hybrid DoA prediction-estimation scheme. Results demonstrate significant improvement compared to the conventional approach.
- 1 May 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 4632-4636
Fast Direction-of-arrival Estimation of Multiple Targets Using Deep Learning and Sparse Arrays
In this work, we focus on improving the Direction-of-Arrival (DoA) estimation of multiple targets/sources from a small number of snapshots. Estimation via the sample covariance matrix is known to perform poorly, since the true manifold structure is not revealed for a small number of samples. First, we explicitly model the sample covariance matrix that is used for the DoA estimation as a noisy version of the true one. Next, we employ a stacked denoising autoencoder (DAE) that predicts a statistically "richer" version of the sampled matrix that is subsequently used for the DoA estimation. Moreover, we consider a limited number of sensors (comparable to the number of sources) in a non-uniform linear configuration and introduce an end-to-end hybrid DoA prediction-estimation scheme. Results demonstrate significant improvement compared to the conventional approach.
- 26 March 2020
- Published by Springer Nature in Wireless Personal Communications
- Vol. 126 (1) , 49-79
Coordinated Caching and QoS-Aware Resource Allocation for Spectrum Sharing
- 26 March 2020
- Published by Springer Nature in Wireless Personal Communications
- Vol. 126 (1) , 49-79
Coordinated Caching and QoS-Aware Resource Allocation for Spectrum Sharing
- Open Access
- 24 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 9 (6) , 770-773
A Stochastic Optimization Approach to Hybrid Processing in Massive MIMO Systems
The high cost and energy consumption of fully digital massive multiple-input multiple-output (MIMO) systems has led to hybrid designs with fewer radio frequency (RF) chains than antennas. In this letter, we propose an efficient hybrid processing algorithm for point-to-point (P2P) massive MIMO systems that operate in either rich or poor scattering environments. The proposed scheme, i.e., hybrid processing via stochastic approximation with Gaussian smoothing (HPSAGS), alternates between a digital baseband and an analog RF precoder/combiner computation step. The method achieves state-of-the-art performance with low computational cost, which is essential for large MIMO systems.
- 24 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 9 (6) , 770-773
Open AccessA Stochastic Optimization Approach to Hybrid Processing in Massive MIMO Systems
The high cost and energy consumption of fully digital massive multiple-input multiple-output (MIMO) systems has led to hybrid designs with fewer radio frequency (RF) chains than antennas. In this letter, we propose an efficient hybrid processing algorithm for point-to-point (P2P) massive MIMO systems that operate in either rich or poor scattering environments. The proposed scheme, i.e., hybrid processing via stochastic approximation with Gaussian smoothing (HPSAGS), alternates between a digital baseband and an analog RF precoder/combiner computation step. The method achieves state-of-the-art performance with low computational cost, which is essential for large MIMO systems.