Abla Bedoui

Description

Dr. Abla Bedoui is a postdoctoral fellow at Long Island University whose interdisciplinary research bridges artificial intelligence, biomedical imaging, and behavioral data analysis. With a foundational Ph.D. in signal processing applied to wireless systems, she has transitioned her expertise to the biomedical domain, where she develops AI-based diagnostic and decision-support tools.

Her work includes deep learning-based segmentation for cardiac MRI and 4D echocardiography, ECG feature analysis for pediatric heart disease diagnosis, and esophageal geometry reconstruction using multi-modal imaging. At LIU, she leads projects in collaboration with the School of Health Professions, where she uses NLP to analyze language and psychological factors linked to behavioral disorders.

Dr. Bedoui is also an experienced educator, having taught courses in artificial intelligence, signal processing, Linux, and programming languages (C, MATLAB). She actively mentors undergraduate and master’s students and has co-supervised thesis work in cardiology and orthopedics. Her teaching style emphasizes hands-on, project-based learning and making complex AI concepts accessible to diverse student groups.

Research

Dr. Bedoui's research is grounded in the application of artificial intelligence across multiple healthcare domains, focusing on three primary data modalities: image data, signal data, and behavioral/psychological data.

In the domain of image data, Dr. Bedoui developed an attention-based deep learning algorithm for accurate segmentation of cardiac structures in MRI. She is currently working on the annotation and analysis of 4D echocardiographic data to facilitate aortic valve segmentation, an essential step in improving diagnostic accuracy for valvular heart diseases. She is also involved in a project applying pattern recognition on knee X-ray images to detect and stage osteoarthritis, offering a non-invasive tool for early diagnosis and disease monitoring. Future work aims to extend these methods to histopathological data for liver-related pathology detection.

For signal data, Dr. Bedoui's research focuses on developing advanced AI algorithms for feature extraction, classification, and biomarker discovery from ECG signals. She goes beyond time-domain analysis by incorporating frequency-domain and higher-order statistical techniques to extract more robust features. She has supervised master’s thesis work that developed convolutional neural networks (CNNs) for cardiovascular disease detection and is currently finalizing a manuscript on deep learning-based discovery of novel ECG biomarkers for atrial fibrillation and related conditions.

In behavioral and psychological data, Dr. Bedoui leads the AI aspect of two interdisciplinary projects with the School of Health Professions. The first uses natural language processing (NLP) to analyze children's speech patterns for developmental assessment through language sampling. The second applies NLP to behavioral analysis to predict factors leading to anger and aggression, aiming to support early intervention strategies and behavioral health management.

At the Center of Excellence, Dr. Bedoui collaborates with a multidisciplinary team of clinicians, engineers, and data scientists to develop and validate virtual twin models of the heart and liver. Her role centers on leveraging patient-specific imaging and signal data to inform these models and ensure their clinical applicability. This collaborative environment has significantly broadened her AI research and translational impact.

Grants and Awards

• Excellence scholarship by the National Center for Scientific and Technical Research (CNRST)
• Erasmus plus grant by the European Union

Selected Publications

• A. Al-Dweik, A. Bedoui and Y. Iraqi, "On the BER Analysis of NOMA Systems," in IEEE Wireless Communications Letters, doi: 10.1109/LWC.2023.3343813.
• A. Bedoui, M. Et-tolba., "A Deep Neural Network-Based Interference Mitigation for MIMO FBMC/OQAM Systems", Frontiers in Communications and Networks, vol.2, pp. 46, 2021.
• A.Bedoui, M. Et-tolba., "A deep learning-based HARQ chase combining for FBMC-OQAM systems, International Journal of Communication Systems, vol.34, Issue.16, 2021
• J R.Pérez , Ó.Fernández, L.Valle, A.Bedoui , M.Et-tolba, R P.Torres, "Experimental Analysis of Concentrated versus Distributed Massive MIMO in an Indoor Cell at 3.5 GHz", Electronics, vol.10, n.14, 2021. Conferences
• Abla Bedoui , Mohamed Et-tolba, Óscar Fernández, Jesús R. Perez , Luis Valle, and Rafael P. Torres, « Performance Analysis of Indoor Distributed Massive MIMO Based on Channel Measurements at 3.5 GHz” Presented in the 16th European conference on antennas and propagation, 27March-01 April 2022, Madrid, Spain.
• A. Bedoui and M. Et-tolba, "A Neuro-Fuzzy based detection approach for HARQ-CC in FBMC/OQAM systems," 2020 9th IFIP International Conference on Performance Evaluation and Modeling in Wireless Networks 10.23919/PEMWN50727.2020.9293073. (PEMWN), 2020, pp. 1-7, doi:
• A. Bedoui, M. Et-tolba and H. Nouasria, "An MMSE Integrated Equalization for HARQ Chase Combining in OQAM-FBMC systems," 2019 15th International Wireless Communications and Mobile Computing Conference (IWCMC), 2019, pp. 811-816, doi: 10.1109/IWCMC.2019.8766473.
• H. Nouasria, M. Et-tolba and A. Bedoui, "New Sensing Matrices Based on Orthogonal Hadamard Matrices For Compressive Sensing," 2019 15th International Wireless Communications and Mobile Computing Conference (IWCMC), 2019, pp. 186-191, doi: 10.1109/IWCMC.2019.8766681.
• A. Bedoui and M. Et-tolba, "A novel Turbo ICI cancellation technique for FBMC/OQAM trough a doubly selective channel," 2018 6th International Conference on Wireless Networks and Mobile Communications (WINCOM), 2018, pp. 1-6, doi: 10.1109/WINCOM.2018.8629757.
• A. Bedoui and M. Et-tolba, "A comparative analysis of filter bank multicarrier (FBMC) as 5G multiplexing technique," 2017 International Conference on Wireless Networks and Mobile Communications (WINCOM), 2017, pp. 1-7, doi: 10.1109/WINCOM.2017.8238200.