Revisiting 2D Foundation Models for Scalable 3D Medical Image Classification
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916051583762432 |
|---|---|
| author | Liu, Han Georgescu, Bogdan Zhang, Yanbo Yoo, Youngjin Baumgartner, Michael Gao, Riqiang Wang, Jianing Zhao, Gengyan Gibson, Eli Comaniciu, Dorin Grbic, Sasa |
| author_facet | Liu, Han Georgescu, Bogdan Zhang, Yanbo Yoo, Youngjin Baumgartner, Michael Gao, Riqiang Wang, Jianing Zhao, Gengyan Gibson, Eli Comaniciu, Dorin Grbic, Sasa |
| contents | 3D medical image classification is essential for modern clinical workflows. Medical foundation models (FMs) have emerged as a promising approach for scaling to new tasks, yet current research suffers from three critical pitfalls: data-regime bias, suboptimal adaptation, and insufficient task coverage. In this paper, we address these pitfalls and introduce AnyMC3D, a scalable 3D classifier adapted from 2D FMs. Our method scales efficiently to new tasks by adding only lightweight plugins (about 1M parameters per task) on top of a single frozen backbone. This versatile framework also supports multi-view inputs, auxiliary pixel-level supervision, and interpretable heatmap generation. We establish a comprehensive benchmark of 12 tasks covering diverse pathologies, anatomies, and modalities, and systematically analyze state-of-the-art 3D classification techniques. Our analysis reveals key insights: (1) effective adaptation is essential to unlock FM potential, (2) general-purpose FMs can match medical-specific FMs if properly adapted, and (3) 2D-based methods surpass 3D architectures for 3D classification. For the first time, we demonstrate the feasibility of achieving state-of-the-art performance across diverse applications using a single scalable framework (including 1st place in the VLM3D challenge), eliminating the need for separate task-specific models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_12887 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Revisiting 2D Foundation Models for Scalable 3D Medical Image Classification Liu, Han Georgescu, Bogdan Zhang, Yanbo Yoo, Youngjin Baumgartner, Michael Gao, Riqiang Wang, Jianing Zhao, Gengyan Gibson, Eli Comaniciu, Dorin Grbic, Sasa Computer Vision and Pattern Recognition 3D medical image classification is essential for modern clinical workflows. Medical foundation models (FMs) have emerged as a promising approach for scaling to new tasks, yet current research suffers from three critical pitfalls: data-regime bias, suboptimal adaptation, and insufficient task coverage. In this paper, we address these pitfalls and introduce AnyMC3D, a scalable 3D classifier adapted from 2D FMs. Our method scales efficiently to new tasks by adding only lightweight plugins (about 1M parameters per task) on top of a single frozen backbone. This versatile framework also supports multi-view inputs, auxiliary pixel-level supervision, and interpretable heatmap generation. We establish a comprehensive benchmark of 12 tasks covering diverse pathologies, anatomies, and modalities, and systematically analyze state-of-the-art 3D classification techniques. Our analysis reveals key insights: (1) effective adaptation is essential to unlock FM potential, (2) general-purpose FMs can match medical-specific FMs if properly adapted, and (3) 2D-based methods surpass 3D architectures for 3D classification. For the first time, we demonstrate the feasibility of achieving state-of-the-art performance across diverse applications using a single scalable framework (including 1st place in the VLM3D challenge), eliminating the need for separate task-specific models. |
| title | Revisiting 2D Foundation Models for Scalable 3D Medical Image Classification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.12887 |