π¬ Iβm currently working at the intersection of Deep Learning and Medical Imaging, developing advanced models for segmentation, classification, and beyond.
π§ My focus includes self-supervised, semi-supervised, and generative learning techniques to tackle real-world challenges where annotated data is scarce or noisy.
ποΈ Iβm also exploring the design and training of foundation models for scalable and generalizable medical image understanding across modalities and tasks.
π€ I'm keen to collaborate on innovative research and translational projects, particularly those with industry relevance and clinical impact.
π MICCAI 2024 Challenge Contributions
π Position | Challenge Name | Title/Method |
---|---|---|
π₯ 1st | AIMS-TBI - Automated Identification of Moderate-Severe Traumatic Brain Injury Lesions | Leveraging Student-Teacher Networks in Self-Supervised Learning for Enhanced TBI Severity Segmentation |
4th | ISLES - Ischemic Stroke Lesion Segmentation | A Two-Stage SSL Approach for Ischemic Stroke Lesion Segmentation |
π₯ 2nd | UWF4DR - Ultra-Widefield Fundus Imaging for Diabetic Retinopathy | Efficient Deep Learning for Ultra-Widefield Fundus Imaging |
π₯ 2nd | FETA - Fetal Tissue Annotation | Pseudo Labeling + 3D Deep Learning Models |
π₯ 3rd | MBH-Seg - Multi-class Brain Hemorrhage Segmentation | Efficient SSL-Based Deep Learning for Hemorrhage Segmentation |
π₯ 3rd | CARE - Real World Medical Image Analysis | Two-Stage SSL for Whole Heart Segmentation in CT and MRI |
4th | CURVAS - Calibration and Uncertainty for Multi-Rater Volume Assessment | xSLTM-UNet Deep Learning Model |
4th | TriALS24 - Triphasic-Aided Liver Lesion Segmentation | SSL-Based Student-Teacher Architecture for Non-Contrast CT |
5th | HNTSMRG - Head and Neck Tumor Segmentation in MR | Self-Supervised xLSTM-UNet for Tumor Segmentation |
5th | MBAS - Multi-class Bi-Atrial Segmentation | Student-Teacher SSL for 3D Bi-Atrial Segmentation |
8th | MARIO - AMD Progression in OCT | Efficient DL Models for Age-Related Macular Degeneration Tracking |
9th | COSAS - Cross-Organ and Scanner Adenocarcinoma Segmentation | Swin-UNet + Parallel Cross-Attention |
9th | TopCoW - Anatomical Segmentation of the Circle of Willis | Pretrained 3D Segmentation Models |
12th | AortaSeg24 - Aortic Branch and Zone Segmentation | SSL-Based Aortic Segmentation with Student-Teacher Architectures |
π§ IEEE ISBI 2023 Challenge Contributions
π Position | Challenge Name |
---|---|
π₯ 1st | CuRIOUS 2022 - Image Registration & Segmentation |
4th | CMRxMotion Challenge |
4th | cSeg-2022 - Infant Cerebellum MRI Segmentation |
6th | ATMβ22 - Multi-Site Multi-Domain Airway Tree Modeling |
9th | NCCT - Intracranial Hemorrhage Segmentation |
10th | ISLES'22 - Ischemic Stroke Lesion Segmentation |
11th | Kidney Parsing Challenge - Renal Cancer Treatment |
16th | Pulmonary Artery Segmentation Challenge |
πΉ MICCAI 2022 Challenge Contributions
π Position | Challenge Name | Method Summary |
---|---|---|
π₯ 1st | CuRIOUS β Correction of Brain Shift with Intra-Operative Ultrasound | Self-Supervised Two-Stage 3D ResUNet |
4th | CMRxMotion Challenge | β |
4th | cSeg-2022 β Infant Cerebellum MRI Segmentation | β |
6th | ATMβ22 β Multi-site, Multi-Domain Airway Tree Modeling | 3D Deep Learning Models |
9th | NCCT β Intracranial Hemorrhage Segmentation | β |
10th | ISLES'22 β Ischemic Stroke Lesion Segmentation | β |
11th | Kidney Parsing Challenge β Renal Cancer Treatment | β |
16th | Pulmonary Artery Segmentation Challenge | β |
π§ MICCAI 2021 Challenge Contributions
π Position | Challenge Name |
---|---|
4th | Diabetic Foot Ulcer Challenge |
4th | FetReg - Placental Vessel Segmentation in Fetoscopy |
5th | Foot Ulcer Segmentation Challenge |
7th | FLARE - Fast and Low GPU Abdominal Organ Segmentation |
10th | Right Ventricular Segmentation in Cardiac MRI |
13th | Feta2021 - Fetal Brain Tissue Segmentation |
6th | Chest XR COVID-19 Detection (Grand Challenge) |
13th | AIROGS - Robust Glaucoma Screening |
5th | KNIGHT Challenge - Kidney Clinical Notes & Imaging Biomarker Discovery |
14th | HECKTOR - Head and Neck Tumor PET/CT Segmentation |
- π« How to reach me: [email protected]