Marawan Elbatel

Marawan Elbatel

Medical Image Analysis

Download my CV here:

About me


I am a first-year PhD student at the Hong Kong University of Science and Technology, working under the supervision of Xiaomeng Li . My PhD is kindly supported by three grants, two of which are from the Hong Kong government. I previously graduated from an Erasmus Mundus MSc in Medical Imaging, fully sponsored by ECEA, under the supervision of Robert Marti.

My main research interests include domain adaptation, representation transfer, and federated learning, with a specific focus on medical imaging applications.

During my Erasmus Mundus MSc, I had the opportunity to study at three universities across three countries, where I was honored to receive five first-authored awards, which included two challenge victories and two best paper awards at MICCAI conference.

Prior to my MSc in Medical Imaging, I obtained my BSc in Computer Engineering from Cairo University. My BSc thesis, under the guidance of AbdelMoneim Bayoumi and with industrial mentorship from DELL Technologies, was centered on object tracking over a multiple-camera network.

Awards

  • HKSAR Government Belt and Road Scholarship (25 awardee in Hong Kong), 2023-2027
  • HKUST PhD Scholarship (RedBird), 2023.
  • MICCAI'23 EndoVis-SynISS Challenge Winners, 2023. (1st Author)
  • MICCAI’22 NVIDIA Breast density federated learning Challenge Winners, 2022. (Joint 1st Author)
  • MICCAI’22 DART Best Paper Award Sponsored by NVIDIA, 2022. (Joint 1st Author)
  • MICCAI’23 PS-FH-AoP Challenge Runner-Up, 2023. (1st author with 0.0002 to a win)
  • MICCAI’23 Tumor Segmentation in 3D Breast Ultrasound Challenge Runner-Up, 2023. (collaboration)
  • Multiple Object Tracking Benchmark, MOT20Det Runner-up during BSc. (1st Author)
  • MICCAI’23 LMIC Participation Grant, 2023. (in person)
  • MICCAI’22 Society Participation Grant (In Person), 2022.
  • Research Fellowship, University of Bradford (one awardee in the department), 2020
  • Research Fellowship, Egypt-Japan Education Partnership (EJEP) (2 awardees in the department), 2019
  • Undergraduate Tuition Fee Scholarship at Cairo University, 2016-2021


Projects




FoPro-KD
Self-prompting Large Vision Models for Few-Shot Medical Image Segmentation

Qi Wu, Yuyao Zhang, and Marawan Elbatel

Undergraduate Research Supervision, 2023

MICCAI'23 Domain Adaptation and Representation Transfer (DART) Workshop, 2023

Paper Code
FoPro-KD
Fourier Prompted Effective Knowledge Distillation for Long-Tailed Medical Image Recognition

Marawan Elbatel, Robert Martí, and Xiaomeng Li

MAIA MSc Thesis, 2023

IEEE Transactions on Medical Imaging (IEEE TMI ), 2023

Paper Code
FedMAS
Federated Model Aggregation via Self-Supervised Priors for Highly Imbalanced Medical Image Classification

Marawan Elbatel, Hualiang Wang, Robert Martí, Huazhu Fu, and Xiaomeng Li

MAIA MSc Thesis, 2023

MICCAI'23 Best Paper Award @ Distributed, Collaborative and Federated Learning (DeCaF) Workshop.

Code
FedMAS
EndoVis: Synthetic data for Instrument Segmentation in Surgery

Marawan Elbatel and Xiaomeng Li

MICCAI'23 Challenge Winners

MICCAI'23 Code
breast FL
ACR-NCI-NVIDIA Federated Learning Challenge

Marawan Elbatel*, Kaouther Mouheb*, and Robert Martí

MICCAI'22 Challenge Winners

University of Girona, Spain

fetoscopy image
SISSI: Seamless Iterative Semi-Supervised Correction of Imperfect Labels

Marawan Elbatel*, Christina Bornberg*, Manasi Kattel, Enrique Almar, Claudio Marrocco, and Alessandro Bria

Cell Cytotoxicity Grading: Object Detection With Noisy and Missing Annotation. Learning from partially and weakly annotated data in a cross-domain setting.

MICCAI'22 Best Paper Award @ Domain Adaptation and Representation Transfer (DART) Workshop.

University of Cassino, Italy

Code
skin diffusion
Real or Fake? Synthetic Skin Lesion Generation with Diffusion Models

Generating synthetic skin lesions images with diffusion models.

University of Girona, Spain.

mammogram image
Breast Cancer Molecular Sub-Types Classification

Working under the supervision of Prof. Robert Marti on the ICEBERG project in the VICOROB lab.

University of Girona, Spain.

ICEBERG
fetoscopy image
Study Buddies: Semi-Supervised Learning

A semi-supervised learning method for placental vessel segmentation

.

Baseline
Atlasb
Brain MRI ATLAS Registration

Using Brain MRI data from registration and atlas construction to position and intensity aware brain tissue segmentation

University of Girona, Spain.

registration
Lung CT Deep Registration

Applying deep learning based registration frameworks (VoxelMorph/SynthMorph/DiffMorph) to Lung CT Exhale/Inhale image registration.

University of Girona, Spain.

eye segmentation
Retinal-Vessel-Segmentation

We implement two methods from DRIVE/Stare images in python. First one is implemented in python originally documented in Farkoo (Matlab). Second one is a hybrid method based on Diaz Et al and Multi-level thresholding with Harris Hawks Optimization.

University of Bourgogne, France, 2021.

Code
Persolo - Person Detection in Crowds

A YOLO-V4 based Pedestrian Detection Model that amplifies the person class bias in YOLO V4 model for more accurate prediction.

Marawan Elbatel

University of Cairo, Egypt, 2020-2021.

MOT20Det Challenge Code
People
Scalable Video Surveillance System for Contact Tracing.

Tracking Multiple People Over a Network of Cameras, Cairo, Egypt.

Dell Technologies, Cairo, Egypt, 2020-2021.

covid diagram
COVID-19 Time series Forecasting.

Working with big-data analytics and deep learning networks to predict and analyze the number of COVID-19 cases in various countries.

Bradford University, UK, 2019-2020.

Timeline

Education & Experience