Researchers have developed a novel two-stage pipeline for simulating abdominal ultrasound images from semantic labels, bypassing the need for patient-specific CT scans during inference. The first stage uses models like the Semantic Diffusion Model (SDM) or Pix2Pix to generate a physics-based image from anatomical segmentations. The second stage refines this image into a realistic ultrasound scan using a segmentation-guided CycleGAN. This approach allows for controlled simulations of both healthy and pathological conditions by editing anatomical maps, though CT-derived data is still required for training. AI
IMPACT Enables more flexible and controllable medical imaging simulations for training and research.
RANK_REASON The item is a research paper detailing a new AI model pipeline for medical image simulation. [lever_c_demoted from research: ic=1 ai=1.0]
- Abdominal Ultrasound With Doppler and Peripheral Hemogram in Assesment Inflammatory Bowel Disease
- child
- computed tomography
- CycleGAN
- Fréchet inception distance
- lpips
- Mae
- Miou-Miou
- Pix2Pix
- Semantic Diffusion Model
- SG-CycleGAN
- Structural Similarity Index Measure
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
