Evidence-Based Text-Conditioned 3D CT Synthesis for Ovarian Cancer

2026-06-27

AI Summary

This study addresses the challenge of limited annotated contrast-enhanced CT data for late-stage ovarian cancer diagnosis, hindered by privacy constraints and scarcity of labeled datasets, which impedes the development of generalizable models. To overcome this, the authors propose OvESyn—the first text-conditioned 3D CT synthesis framework tailored for abdominopelvic tumors—that generates realistic scans using only automated segmentation masks and routine preoperative clinical metadata, without requiring original radiology reports. Built upon a latent diffusion model, OvESyn integrates vision-language alignment and generator domain adaptation techniques, trained on data from 493 high-grade serous ovarian cancer patients. Experiments demonstrate that the full OvESyn model achieves superior fidelity in distribution and intensity (FID2.5D = 29.35, Precision = 0.671), while a generator variant attains a coverage of 0.645, effectively revealing the inherent trade-off between fidelity and diversity.

Abstract

Ovarian cancer is frequently diagnosed at an advanced stage, making preoperative contrast-enhanced computed tomography (CT) central to staging and surgical planning; yet the scarcity of annotated imaging data, compounded by privacy regulations, limits the development of generalizable computational models in this domain. Text-conditioned 3D CT synthesis has shown promise, but existing pipelines depend on paired radiology reports and have been evaluated only on chest CT. We propose OvESyn (Ovarian Evidence-based Synthesis), a framework that constructs standardized Findings and Impression sections directly from CT-derived imaging descriptors and routine clinical metadata, without any original radiology report, and uses them to condition a latent diffusion model adapted to 493 high-grade serous ovarian carcinoma patients. This is the first text-conditioned 3D CT synthesis framework adapted to an abdomino-pelvic oncologic setting. A systematic ablation over two adaptation axes, vision-language encoder alignment and generator fine-tuning, identifies generator domain adaptation as the operative mechanism for crossing the domain gap and establishing the target anatomy: without it, synthesis remains anchored to the thoracic pretraining domain, with Precision and Recall collapsing to zero and FID2.5D exceeding 140, regardless of encoder alignment. Encoder alignment instead refines intensity and fine detail. The full OvESyn attains the best distributional and intensity fidelity (FID2.5D 29.35, Precision 0.671, Wasserstein-1 0.044), while the generator-only variant maximizes coverage (Recall 0.645), reflecting a fidelity/coverage trade-off governed by encoder adaptation. Requiring only automatic segmentations and routine preoperative metadata, OvESyn supports transferability to report-scarce settings and provides a foundation for synthetic cohort generation in abdomino-pelvic oncologic imaging.

Problem

• ovarian cancer
• 3D CT synthesis
• text-conditioned generation
• data scarcity
• abdomino-pelvic imaging

Innovation

• text-conditioned synthesis
• latent diffusion model
• domain adaptation
• ovarian cancer imaging
• 3D CT generation