MAISI-v2: Accelerated 3D High-Resolution Medical Image Synthesis with Rectified Flow and Region-specific Contrastive Loss

TL;DR

MAISI-v2 is a novel accelerated 3D medical image synthesis framework that combines rectified flow and region-specific contrastive loss to produce high-quality images efficiently and with improved condition fidelity.

Contribution

It introduces the first accelerated 3D medical image synthesis method integrating rectified flow and a novel contrastive loss for better condition consistency.

Findings

Achieves 33x faster inference with state-of-the-art image quality.

Demonstrates effective use for data augmentation in downstream tasks.

Provides open-source code and tools for reproducibility.

Abstract

Medical image synthesis is an important topic for both clinical and research applications. Recently, diffusion models have become a leading approach in this area. Despite their strengths, many existing methods struggle with (1) limited generalizability that only work for specific body regions or voxel spacings, (2) slow inference, which is a common issue for diffusion models, and (3) weak alignment with input conditions, which is a critical issue for medical imaging. MAISI, a previously proposed framework, addresses generalizability issues but still suffers from slow inference and limited condition consistency. In this work, we present MAISI-v2, the first accelerated 3D medical image synthesis framework that integrates rectified flow to enable fast and high quality generation. To further enhance condition fidelity, we introduce a novel region-specific contrastive loss to enhance the sensitivity to region of interest. Our experiments show that MAISI-v2 can achieve SOTA image quality with $33 \times$ acceleration for latent diffusion model. We also conducted a downstream segmentation experiment to show that the synthetic images can be used for data augmentation. We release our code, training details, model weights, and a GUI demo to facilitate reproducibility and promote further development within the community.