The Pattern Recognition Lab at MICCAI 2025
The Pattern Recognition Lab is proud to announce its strong and diverse presence at MICCAI 2025 in Daejeon, South Korea, from September 23 to 27. MICCAI (28th International Conference on Medical Image Computing and Computer Assisted Intervention) is a leading venue for research at the intersection of medical imaging, machine learning, and clinical applications.
This year, the Pattern Recognition Lab will contribute to every part of the program. Our researchers will present novel methods and foundational advances at the main conference. Additionally, the lab is involved in specialized workshops that highlight emerging trends and focused research areas. Several teams have also successfully competed in international challenges that benchmark new approaches to critical clinical tasks. Further contributions will be presented at CLINICCAI, which emphasizes the clinical translation of computational methods, and at the Open Data Track, where our researchers will share datasets and tools to promote reproducible science. Finally, our doctoral researchers will present their projects at the PhD consortium, an international forum for early-stage research discussions.
Together, these contributions highlight the breadth of topics covered by the Pattern Recognition Lab and the collaborative strengths of our team, ranging from theoretical innovations to clinical and translational applications.
🔹 Main Conference
- Joshua Scheuplein – DINO Adapted to X-Ray (DAX): Foundation Models for Intraoperative X-Ray Imaging (Poster, Sept 26, 15:30–17:00)
🔗 Paper - Farid Tasharofi – FIND-Net: Fourier-Integrated Network with Dictionary Kernels for Metal Artifact Reduction (Poster Session 2, Sept 25, B102)
🔗 Paper - Rajesh Madhipati, Sheethal Bhat – CXR-CML: Improved Zero-Shot Classification of Long-Tailed Multi-Label Diseases in Chest X-Rays (Poster Session 3, Sept 25)
🔗 Paper - Sheethal Bhat – Exemplar Med-DETR: Toward Generalized and Robust Lesion Detection in Mammogram Images and Beyond (Poster Session 3, Sept 25)
🔗 Paper
🔹 Workshops
- Yipeng Sun – LSTT: Latent Spatio-Temporal Transformer for Non-Rigid Motion Compensation in CBCT (RIME Workshop, Sept 27, Oral)
- Eduardo Castañeda – Fast Multi-Label Parameterization of the Left Atrium by Learned Template Morphing (STACOM, Sept 27)
- Linda Vorberg – Leveraging Last-Known-Well Times for Radiomics-Based Stroke Onset Estimation from Non-Contrast CT (SWITCH Workshop, Sept 23, 14:15 Talk)
- Karim Elbabary – MM-DETR: Emulating the Diagnostic Clinical Workflow in Multi-View Multi-Modal Mammography Mass Detection (Deep-Breath, Sept 23, Poster 9:55–11:15)
- Badhan Das – VIViT: Variable-Input Vision Transformer Framework for 3D MR-Image Segmentation (MLMI Workshop, Oral)
🔹 Challenges
- Luis Rivera, Martin Mayr – BraTS 2025, Task 10: Assessing the Heterogeneous Histologic Landscape of Glioma (Top-3 Method, Oral, Sept 23, 8:15) #BraTS2025
- Siyuan Mei – SynthRAD 2025:
• 1st place, Task 2 (CBCT-to-CT synthesis, Oral Sept 23, 17:15)
• 4th place, Task 1 (MRI-to-CT synthesis, Poster Sept 23, 15:20) - Dr. Tri-Thien Nguyen – ODELIA Grand Challenge: Multi-Sequence Attention-Based MIL and Focal Loss Optimization (DeepBreath Workshop, Sept 23, Oral)
🔹 CLINICCAI
- Dr. Tri-Thien Nguyen – Retrospective Analysis of Foundation Models for Breast MR Image Analysis: Comparing Contrast-Enhanced vs Non-Contrast Protocol (Oral, Sept 25, 13:30)
🔹 Open Data Track
- Dr. Frauke Wilm – A COCO-Formatted Instance-Level Dataset for Plasmodium Falciparum Detection in Giemsa-Stained Blood Smears (Oral, Sept 26, 13:45–14:00)
🔗 Preprint
🔹 PhD Consortium
- Luis Rivera – Interconnected Pathology: A Graph-Based Framework for Analyzing Multimodal Histopathology (Poster, Sept 24, 15:30–17:00, DCC 2)
- Sheethal Bhat – Improving Clinical Usability of Deep Learning for Radiographic Diagnosis: Learning under Constraints of Specificity, Data, and Annotation (Poster, Sept 24, 15:30–17:00, DCC 2)
We thank the MICCAI Society, the Student Board, the workshop organizers, and the MICCAI 2025 organizing committee for their efforts in making this event possible. We are also grateful to our collaborators and research partners for their continued support. The Pattern Recognition Lab looks forward to engaging with colleagues, exchanging ideas, and building new collaborations during and after the conference.
