NV-Generate-Body: 3D Latent Diffusion Model forControllable Anatomy, Tumor, and Demographic Body Mask Synthesis

Key points

• Curating large labeled CT datasets is expensive and constrained by patient privacy.1
• The model operates in the compressed latent space of a frozen mask autoencoder, generating $256^3$ volumetric label maps conditioned on a 19-dimensional vector via Adaptive Group Normalization and cross-attention.1
• A size-adaptive ROI loss with spatial halo weighting and a paired tumor injection pipeline enable accurate tumor size conditioning despite the scarcity of annotated tumor cases.1
• Trained on 8,571 masks across 26 cohorts spanning 132 labels, the model achieves Spearman $ρ= 0.70$--$0.92$ for organ size, 85--100% tumor recall for lung, pancreatic, colon, and bone lesions, and strong demographic alignment (BMI $ρ= 0.78$, sex-conditioned perfect prostate control).1

Sources (1)

• [1]NV-Generate-Body: 3D Latent Diffusion Model forControllable Anatomy, Tumor, and Demographic Body Mask Synthesis
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 07:26 PM

  We present NV-Generate-BodyMask, a 3D latent diffusion model that synthesizes whole-body CT segmentation masks with precise, joint control over organ anatomy, tumor burden, and patient demographics.

  Curating large labeled CT datasets is expensive and constrained by patient privacy.

Extractive summary: sentences quoted from the sources.