Improving realism in abdominal ultrasound simulation combining a segmentation‐guided loss and polar coordinates training
Our segmentation-guided, polar-coordinates-trained CycleGAN framework significantly reduces hallucinations, ensuring anatomical consistency, and realism in simulated abdominal US images, surpassing existing methods.
Authors
- Ignacio Larrabide (ORCID: https://orcid.org/0000-0002-7745-1964)
- Emmanuel Iarussi (ORCID: https://orcid.org/0000-0001-7438-9299)
- José Ignacio Orlando (ORCID: https://orcid.org/0000-0001-9734-5571)
- Alejandro Díaz (ORCID: https://orcid.org/0000-0001-7536-3269)
- Santiago Vitale (ORCID: https://orcid.org/0000-0003-4973-5342)
Institutions
- Universidad Torcuato Di Tella (AR)
- Universidad Nacional del Centro de la Provincia de Buenos Aires (AR)
- Centro Científico Tecnológico - San Juan (AR)
- Centro Científico Tecnológico - Tandil (AR)
Publication Details
- Journal
- Medical Physics
- Published
- 2025-03-30
- DOI
- https://doi.org/10.1002/mp.17801
- Citations
- 2
- Primary Topic
- Generative Adversarial Networks and Image Synthesis
- Type
- article
- Field-Weighted Citation Impact
- 2.55
Funders
- Consejo Nacional de Investigaciones Científicas y Técnicas
- Agencia Nacional de Promoción de la Investigación, el Desarrollo Tecnológico y la Innovación