Label-Preserving Fourier Domain Adaptation with Generative-AI and Real Style Donors for 3D Mitral Valve Segmentation
Organizations: Department of Computer and Software Engineering, College of Electrical and Mechanical Engineering (CEME), NUST, Rawalpindi, Pakistan · Electrical Engineering Department, Namal University, Mianwali, Pakistan · Electrical Engineering Department, Information Technology University (ITU), Lahore, Pakistan
Abstract
Accurate segmentation of the mitral valve from three-dimensional transesophageal echocardiography (3D TEE) is severely restricted by the scarcity of publicly available expert- annotated volumetric datasets. To address this critical limitation, we investigated Fourier Domain Adaptation (FDA), utilizing general-purpose generative-AI images (from platforms like Chat- GPT, Google Gemini, and Microsoft MAI) as annotation-free style donors for 3D mitral valve segmentation. We employed a DeepMitral residual U-Net model on the public MVSeg2023 dataset, comparing six conditions over three runs: real-only training, traditional augmentation, FDA with each generative source, and a non-AI FDA control using real-image style donors, plus an optimization-matched real-only control and a \b{eta} sensitivity analysis. Case-level paired testing showed that FDA with these generative sources led to significant reductions in boundary- and surface-related errors across all three donor conditions relative to the primary real-only baseline. However, the impact on the core segmentation metric, Generalized Dice Score (GDS), proved generator-dependent. ChatGPT-FDA showed a significant GDS advantage over the optimization-matched control but not over the primary baseline, and Gemini-FDA showed no significant GDS difference. Notably, only MAI-FDA provided a statistically robust and significant increase in GDS compared to the primary baseline. The \b{eta} sensitivity was not robust across training runs. These findings confirm that FDA remains an important augmentation mechanism. While generative-AI images offer a practical means of acquiring style diversity without new annotations, the non-AI control performed comparably to ChatGPT-FDA, so we do not conclude that this approach is superior to training with real-image donors.
Figures & tables
| Condition | GDS | HD95 (mm) | Surface Distance (mm) |
|---|---|---|---|
| Real-only baseline | 0.8198 0.0117 | 6.0005 1.6528 | 1.2911 0.2886 |
| Traditional augmentation | 0.8077 0.0175 | 6.2557 2.4475 | 1.3235 0.6363 |
| Real + ChatGPT-FDA | 0.8268 0.0307 | 3.0569 0.1696 | 0.5866 0.1098 |
| Real + Gemini-FDA | 0.8192 0.0291 | 3.0816 0.5196 | 0.5712 0.1460 |
| Real + MAI-FDA | 0.8309 0.0129 | 3.6890 1.0765 | 0.7028 0.2258 |
| Real + Non-AI FDA | 0.8259 0.0268 | 3.0208 0.3519 | 0.5824 0.0765 |
| Reference | Condition | GDS ( ) | HD95, mm ( ) | Surface Dist., mm ( ) |
|---|---|---|---|---|
| 390-update real-only | ChatGPT-FDA | ( ) | ( ) | ( ) |
| 390-update real-only | Gemini-FDA | ( ) | ( ) | ( ) |
| 390-update real-only | MAI-FDA | ( ) | ( ) | ( ) |
| 540-update real-only | ChatGPT-FDA | ( ) | ( ) | ( ) |
| 540-update real-only | Non-AI FDA | ( ) | ( ) | — ( ) |
| GDS | HD95 (mm) | Surface Dist. (mm) | |
|---|---|---|---|
| 0.005 | |||
| 0.010 | |||
| 0.020 |
Medical Image Segmentation
Medical image segmentation delineates organs, lesions, and other structures in scans to support measurement, diagnosis, treatment planning, and research.
See all 839 papers