# Rad-JEPA 3D: Radiology Joint-Embedding Predictive Model for 3D Computed Tomography

arXiv: https://arxiv.org/abs/2607.26196 · HTML: https://mlsanity.com/paper/2607.26196/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Mamba for medical image analysis, one of 10 directions in State-space sequence models (S4, Mamba).
- Builds on 37 of its 40 references that are in the map.
- First posted on arXiv 28 Jul 2026.

## Placed in

- State-space sequence models (S4, Mamba) / [Mamba for medical image analysis](https://mlsanity.com/ssm/mamba-for-medical-image-analysis/)

## Builds on (most cited first, 30 of 37)

- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Bootstrap your own latent: A new approach to self-supervised Learning](https://mlsanity.com/paper/2006.07733/) (2020)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Barlow Twins: Self-Supervised Learning via Redundancy Reduction](https://mlsanity.com/paper/2103.03230/) (2021)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints](https://mlsanity.com/paper/2305.13245/) (2023)
- [Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality](https://mlsanity.com/paper/2405.21060/) (2024)
- [LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day](https://mlsanity.com/paper/2306.00890/) (2023)
- [Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture](https://mlsanity.com/paper/2301.08243/) (2023)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [Revisiting Feature Prediction for Learning Visual Representations from Video](https://mlsanity.com/paper/2404.08471/) (2024)
- [Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image Analysis](https://mlsanity.com/paper/2111.14791/) (2021)
- [Big Self-Supervised Models Advance Medical Image Classification](https://mlsanity.com/paper/2101.05224/) (2021)
- [M3D: Advancing 3D Medical Image Analysis with Multi-Modal Large Language Models](https://mlsanity.com/paper/2404.00578/) (2024)
- [Exploring scalable medical image encoders beyond text supervision](https://mlsanity.com/paper/2401.10815/) (2024)
- [V-JEPA 2.1: Unlocking Dense Features in Video Self-Supervised Learning](https://mlsanity.com/paper/2603.14482/) (2026)
- [HealthGPT: A Medical Large Vision-Language Model for Unifying Comprehension and Generation via Heterogeneous Knowledge Adaptation](https://mlsanity.com/paper/2502.09838/) (2025)
- [VoCo: A Simple-yet-Effective Volume Contrastive Learning Framework for 3D Medical Image Analysis](https://mlsanity.com/paper/2402.17300/) (2024)
- [Brain-JEPA: Brain Dynamics Foundation Model with Gradient Positioning and Spatiotemporal Masking](https://mlsanity.com/paper/2409.19407/) (2024)
- [Med3DVLM: An Efficient Vision-Language Model for 3D Medical Image Analysis](https://mlsanity.com/paper/2503.20047/) (2025)
- [A generalizable 3D framework and model for self-supervised learning in medical imaging](https://mlsanity.com/paper/2501.11755/) (2025)
- [Models Genesis](https://mlsanity.com/paper/2004.07882/) (2020)
- [Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging](https://mlsanity.com/paper/2510.20639/) (2025)

## Built on by (most cited first, 1 of 1)

- [The Predictability Bias of Joint-Embedding Predictive Learning in Medical Imaging](https://mlsanity.com/paper/2610.13237/) (2026)

## Abstract (arXiv)

Self-supervised pretraining is central to 3D medical image analysis, where unlabeled CT volumes are abundant but expert annotations are scarce. Yet existing volumetric encoders often fail to preserve the coarse spatial and geometric structure that downstream reasoning depends on, limiting their performance on organ disentanglement, abnormality detection, and spatial understanding when paired with language models. We introduce Rad-JEPA 3D, a joint-embedding predictive framework that learns volumetric CT representations by predicting the latent features of a complete scan from a masked view. At its core is a hybrid H-Mamba encoder that fuses a Mamba state-space branch, which models inter-slice continuity through sequential scanning, with a grouped-query attention branch, which captures cross-plane spatial context, combined through a lightweight per-token router. To improve the quality of intermediate representations, we further propose Hidden States Orthogonal Regularization (HSOR), which aligns student-teacher hidden states and reduces feature redundancy throughout the encoder. This layer-wise regularization produces more consistent and discriminative volumetric representations, leading to improved performance on organ recognition and spatial reasoning tasks. Pretrained on approximately 120,000 CT scans, Rad-JEPA 3D attains state-of-the-art results despite its compact size: with only 4.0B total parameters, it achieves competitive results with state-of-the-art on closed-ended VQA and the best average spatial-reasoning score on the Spatial-Med benchmark. Ablation studies confirm that the hybrid block and HSOR contribute complementary gains, and that the induced spatial structure can substitute for raw language-model scale on volumetric reasoning tasks.