Dorina Thanou
EPFL STI IEM LTS4
ELD 241 (Bâtiment ELD)
Station 11
1015 Lausanne
+41 21 693 75 91
+41 21 693 47 09
Office:
ELE 239
EPFL
>
STI
>
IEM
>
LTS4
Web site: Web site: https://lts4www.epfl.ch/
EPFL SB MATH MDS1
MA C2 543 (Bâtiment MA)
Station 8
1015 Lausanne
Web site: Web site: https://mds.epfl.ch/
EPFL > VPA > AICENTER > AICENTER-GE
Fields of expertise
Data science, Machine learning, Signal processing, AI for medicine and biology
Biography
Dorina Thanou is a senior researcher and lecturer at EPFL, leading strategic research initiatives on AI for Health and Science under the EPFL AI Center. Her research interests include graph-based signal processing for data representation and analysis, as well as machine learning, with a particular focus on the design of interpretable and robust models for biomedicine. She is currently leading many interdisciplinary collaborations that aim at fostering the adoption of AI in medicine and health.Dorina got her M.Sc. (August 2010) and Ph.D. (February 2016) in Communication Systems and Electrical Engineering respectively, both from EPFL, Switzerland, and her Diploma in Electrical and Computer Engineering (July 2008) from the University of Patras, Greece. The topic of her PhD was on representation learning algorithms for graph structured data. In summer 2014, she was a research intern with Microsoft Research, Redmond, USA, working on the compression of 3D point clouds. Between 2016 and 2020, Dorina has been with the Swiss Data Science Center, where she mainly worked on many interdisciplinary research collaborations, in academia and industry. Between 2020-2024, she was with the Center for Intelligent Systems at EPFL, leading the Intelligent Systems for Medicine and Health research pillar.
Among other technical activities, Dorina is currently one of the co-organizers of the Data Science on Graphs webinar series, initiated by the Data Science Initiative of the IEEE Signal Processing Society. She has been the Finance chair for the IEEE Data Science Workshop in 2018, the Finance chair and the invited speakers co-chair for the Graph Signal Processing workshop in 2018. She has served as an Associate Editor of the IEEE Transactions on Image Processing (2021-2022), IEEE Transactions on Signal and Information Processing over Networks (2022-), an area chair for the Learning on Graphs conference, and a reviewer for many international conferences and journals in signal processing and machine learning, such as IEEE Transactions on Signal Processing, IEEE ICASSP, NeurIPS, ICLR among others. She is currently a member of the EURASIP technical area committee on Biomedical Images & Signal Analytics (BISA), and a steering committee member of the Data Science Initiative of the IEEE Signal Processing Society.
Dorina received the Best Student Paper Award at ICASSP 2015, and the Best Paper Award at PCS 2016. In October 2021, she has been an elected an ELLIS Scholar. She is an IEEE senior member.
Awards
ELLIS Scholar in Geometric Deep Learning
2021
Best Student Paper Award, IEEE International Conference on Speech and Signal Processing (ICASSP).
2015
Top 10% Paper Award, IEEE International Conference in Image Processing (ICIP) .
2015
Best Paper Award, Picture Coding Symposium (PCS)
2016
IEEE Senior Member
2023
Publications
Infoscience publications
Publications
Journal Articles
Towards AI-assisted cardiology: a reflection on the performance and limitations of using large language models in clinical decision-making
Eurointervention. 2023-12-01. DOI : 10.4244/EIJ-D-23-00461.Predictive performance of multi-model ensemble forecasts of COVID-19 across European nations
Elife. 2023-04-21. DOI : 10.7554/eLife.81916.Graph Signal Separation Based on Smoothness or Sparsity in the Frequency Domain
IEEE Transactions On Signal And Information Processing Over Networks. 2023-01-01. DOI : 10.1109/TSIPN.2023.3254443.Deep learning-based prediction of future myocardial infarction using invasive coronary angiography: a feasibility study
Open Heart. 2023-01-01. DOI : 10.1136/openhrt-2022-002237.Trend estimation and short-term forecasting of COVID-19 cases and deaths worldwide
Proceedings Of The National Academy Of Sciences Of The United States Of America. 2022-08-09. DOI : 10.1073/pnas.2112656119.Reconstruction of Time-Varying Graph Signals via Sobolev Smoothness
Ieee Transactions On Signal And Information Processing Over Networks. 2022-01-01. DOI : 10.1109/TSIPN.2022.3156886.Combining anatomical and functional networks for neuropathology identification: A case study on autism spectrum disorder
Medical Image Analysis. 2021-04-01. DOI : 10.1016/j.media.2021.101986.node2coords: Graph Representation Learning with Wasserstein Barycenters
Ieee Transactions On Signal And Information Processing Over Networks. 2021-01-01. DOI : 10.1109/TSIPN.2020.3041940.Graph Signal Processing for Machine Learning: A Review and New Perspectives
Ieee Signal Processing Magazine. 2020-11-01. DOI : 10.1109/MSP.2020.3014591.Mask Combination of Multi-Layer Graphs for Global Structure Inference
Ieee Transactions On Signal And Information Processing Over Networks. 2020-01-01. DOI : 10.1109/TSIPN.2020.2995515.Learning Graphs From Data: A Signal Representation Perspective
IEEE Signal Processing Magazine. 2019-05-01. DOI : 10.1109/MSP.2018.2887284.Graph-based Transform Coding with Application to Image Compression
IEEE Transactions on Image Processing. 2019. DOI : 10.1109/TIP.2019.2932853.Learning of robust spectral graph dictionaries for distributed processing
Eurasip Journal On Advances In Signal Processing. 2018-10-24. DOI : 10.1186/s13634-018-0584-2.Learning heat diffusion graphs
IEEE Transactions on Signal and Information Processing over Networks. 2017. DOI : 10.1109/Tsipn.2017.2731164.Graph-based compression of dynamic 3D point cloud sequences
IEEE Transactions on Image Processing. 2016. DOI : 10.1109/Tip.2016.2529506.Learning Laplacian Matrix in Smooth Graph Signal Representations
IEEE Transactions on Signal Processing. 2016. DOI : 10.1109/TSP.2016.2602809.Learning Parametric Dictionaries for Signals on Graphs
IEEE Transactions on Signal Processing. 2014. DOI : 10.1109/Tsp.2014.2332441.Distributed average consensus with quantization refinement
IEEE Transactions on Signal Processing. 2013. DOI : 10.1109/Tsp.2012.2223692.Conference Papers
Tertiary Lymphoid Structures Generation Through Graph-Based Diffusion
2024-01-01. 5th International Workshop on Graphs in Biomedical Image Analysis (GRAIL) / Workshop on Overlapped Celi on Tissue - Cell Detection from Cell-Tissue Interaction Challenge (OCELOT) / 5th MICCAI Workshop / 1st MICCAI Challenge Workshop, Vancouver, CANADA, SEP 23-OCT 04, 2023. p. 37-53. DOI : 10.1007/978-3-031-55088-1_4.Anatomy-informed multimodal learning for myocardial infarction prediction
2022-12-02. MedNeurIPS.Attention-based learning of views fusion applied to myocardial infarction diagnosis from x-ray CT
2022-12-02. MedNeurIPS.Predicting future myocardial infarction from angiographies with deep learning
2021. Medical Imaging meets NeurIPS 2021, [Online only], December 14, 2021.Interpretable Stability Bounds for Spectral Graph Filters
2021-01-01. International Conference on Machine Learning (ICML), ELECTR NETWORK, Jul 18-24, 2021.A Graph Signal Processing Framework for the Classification of Temporal Brain Data
2020-01-01. 28th European Signal Processing Conference (EUSIPCO), ELECTR NETWORK, Jan 18-22, 2021. p. 1180-1184. DOI : 10.23919/Eusipco47968.2020.9287486.Height And Weight Estimation From Unconstrained Images
2020-01-01. IEEE International Conference on Acoustics, Speech, and Signal Processing, Barcelona, SPAIN, May 04-08, 2020. p. 2298-2302. DOI : 10.1109/ICASSP40776.2020.9053363.Learning sparse models of diffusive graph signals
2017. ESANN, April.Graph learning under sparsity priors
2017. International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, USA, March 5-9, 2017. p. 6523-6527. DOI : 10.1109/ICASSP.2017.7953413.Learning time varying graphs
2017. International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, USA. p. 2826-2830. DOI : 10.1109/ICASSP.2017.7952672.Graph Transform Learning for Image Compression
2016. Picture Coding Symposium (PCS), Nuremberg, Germany. DOI : 10.1109/PCS.2016.7906368.Distributed Signal Processing with Graph Spectral Dictionaries
2015. Allerton Conference on Communication, Control, and Computing. p. 1391-1398. DOI : 10.1109/ALLERTON.2015.7447171.Graph-based motion estimation and compression for dynamic 3D point cloud compression
2015. IEEE International Conference on Image Processing (ICIP), Quebec city, Canada, September, 2015. p. 3235-3239. DOI : 10.1109/ICIP.2015.7351401.Multi-Graph Learning of Spectral Graph Dictionaries
2015. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brisbane, Australia, April, 2015. p. 3397-3401. DOI : 10.1109/ICASSP.2015.7178601.Laplacian Matrix Learning for Smooth Graph Signal Representation
2015. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brisbane, Australia, April, 2015. p. 3736-3740. DOI : 10.1109/ICASSP.2015.7178669.Parametric dictionary learning for graph signals
2013. IEEE Global Conference on Signal and Information Processing (GlobalSIP), Austin, Texas, December, 2013. p. 487-490. DOI : 10.1109/GlobalSIP.2013.6736921.Progressive quantization in distributed average consensus
2012. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Kyoto, Japan, March, 2012. p. 2677-2680. DOI : 10.1109/ICASSP.2012.6288468.Compressed classification of observation sets with linear subspace embeddings
2011. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Prague, Czech Republic, May 22-27, 2011. p. 1353-1356. DOI : 10.1109/ICASSP.2011.5946663.Comparison of time and frequency domain interpolation implementations for MB-OFDM UWB transmitters
2010. 2010 IEEE International Symposium on Circuits and Systems (ISCAS 2010), Paris, France, May 30 -June 2 2010. p. 2143-2146. DOI : 10.1109/ISCAS.2010.5536948.Polynomial Filter Design for Quantized Consensus
2010. European Signal Processing Conference (EUSIPCO), Aalborg, Denmark, August 23-27, 2010. p. 184-188.Theses
Novel Methods For Detection And Analysis Of Atypical Aspects In Speech
Lausanne, EPFL, 2023. DOI : 10.5075/epfl-thesis-9785.Graph Signal Processing
Lausanne, EPFL, 2016. DOI : 10.5075/epfl-thesis-6925.Reports
Functionalized nanofiber-enhanced filter media for fine particles and heavy metals removal in flue gases
2015Progressive quantization in distributed average consensus
2011Patents
Methods and apparatuses for encoding and decoding digital images or video streams
US11122298 ; CN110024391 ; EP3549344 ; US2020228840 ; EP3549344 ; CN110024391 ; WO2018100503 ; IT201600122898 . 2018.Student Projects
Biofuels vs food
2008Teaching & PhD
Teaching
Electrical and Electronics Engineering
Courses
Network machine learning
Fundamentals, methods, algorithms and applications of network machine learning and graph neural networks
Graph representations for biology and medicine
Systems of interacting entities, modeled as graphs, are pervasive in biology and medicine. The class will cover advanced topics in signal processing and machine learning on graphs and networks, and will showcase applications of the tools in biomedicine.