CS Paper Radar

Technically Plausible but Clinically Misleading? Expert Evaluation of Patient-Personalized Synthetic Prostate MRI

Gabriel Paulo Maglalang Israel, Sol Gedde, Alvaro Fernandez-Quilez

eess.IV ·submitted 2026-08-17 06:37 UTC · v1
published

Abstract

Magnetic resonance imaging (MRI) is central to prostate cancer assessment, yet its acquisition is costly and time-consuming, making it a major bottleneck in the patient care pathway. Personalized MRI synthesis is a promising direction because it may allow the generation of clinically realistic images tailored to individual patients while reducing the dependence on scanner-acquired imaging. In this work, we in-vestigate whether patient-personalized MRI synthesis produces images that experts perceive as clinically plausible and how such personaliza-tion influences expert interpretation. We synthesize patient-personalized prostate MRI using a 3D diffusion model conditioned on routine pre-imaging clinical variables age and PSA. Using a large publicly available dataset, we first verify that synthesized images are technically compara-ble to scanner-acquired MRIs using standard image similarity metrics. We then conduct a pilot, single-blinded, randomized expert study on a subset of cases sampled to reflect variation in patient clinical pro-files. A radiologist compared personalized and unconditioned synthetic MRIs using paired assessments and Likert-scale ratings of clinical plausibility, confidence, and risk of misleading interpretation, supplemented by qualitative feedback. Although personalized synthetic MRIs appeared technically plausible, expert interpretation highlighted variability in con-fidence and potential risks of misleading cues when routine clinical data was embedded into image generation. These findings suggest that while personalized synthesis may be technically feasible, careful assessment is needed to understand how generated cues influence clinician interpreta-tion before such systems can be safely integrated into clinical workflows.

Comment
To appear in SCAI 2026 proceedings

Closest papers

Nearest neighbours by meaning across the whole index.

Conditional Diffusion-Based Generation of T2-Weighted Prostate MRI for Clinically Significant Prostate Cancer Classification

cs.CV ·2026-11-01 ·Xavier Rafael-Palou, Alejandro Vergara-Richart, Ignacio Iborra Roncales, Carmen Prieto-de-la-Lastra, Andreu Antolin-Redondo +2 more
similarity 0.71
Abstract

Scarce and unrepresentative training data limit the performance of prostate MRI prediction models. Synthetic T2-weighted (T2W) MRI may augment training data containing clinically significant prostate cancer (csPCa) and transition zone (TZ) lesions. In this retrospective, single-center, ethics-committee-approved study, we developed a conditional generative model for T2W prostate MRI extending MAISI-v2 with csPCa class, zonal, and lesion-mask-guided conditioning, incorporating a relative advantage weighted loss fine-tuned on 1,303 patients (1,654 lesions). Fidelity was assessed by incremental ablation, and downstream utility was evaluated with T2W-based csPCa classifiers across five simulated training-data scenarios, each testing 18 synthetic-data configurations. The proposed model reduced Fréchet inception distance (FID) from 13.28 (pretrained MAISI-v2) to $2.04 \pm 0.06$ internally (PI-CAI: $3.83 \pm 0.51$). Synthetic augmentation significantly increased AUC in all five scenarios (+0.027 to +0.088; adjusted $p = 0.016$--$0.043$), though cross-institutional augmentation of PI-CAI yielded a non-significant gain (+0.013; $p = 0.397$). Overall, our model improved T2W MRI fidelity, and synthetic images improved csPCa classification under simulated data constraints, although benefit did not transfer significantly across institutions.

0.21

Diffusion-Based Synthesis of Complete 3D Biparametric Prostate MRI

cs.CV ·2026-10-22 ·Claudia Giardina, Montse Pardàs, Verónica Vilaplana
publishedhot direction similarity 0.71
Abstract · Presented at the 6th Deep Generative Models Workshop (DGM4MICCAI), MICCAI 2026

Synthetic biparametric MRI (bpMRI) may help alleviate the limited availability of large and representative prostate MRI datasets for deep learning applications. In this work, we present a conditional 3D diffusion framework for anatomy-conditioned synthesis of complete prostate bpMRI from anatomical masks. The framework supports sequential synthesis, in which T2w and diffusion modalities are generated by separately trained models, and joint synthesis, in which all bpMRI modalities are generated simultaneously within a single model. Five configurations were trained using PI-CAI, FastMRI Prostate, and PROSTATEx datasets to investigate different conditioning schemes, target modalities, and synthesis strategies. Evaluation using distribution-based metrics, paired-image similarity, and qualitative assessment demonstrated the feasibility of both synthesis pathways. Joint synthesis achieved competitive distributional fidelity while generating complete bpMRI, whereas direct synthesis of ADC and computed high-b-value DWI outperformed their derivation from synthetic diffusion-weighted images.

0.34

CONTEXT-PCa: Evaluating the role of clinical context in MRI-based AI for detecting clinically significant prostate cancer (study protocol)

cs.AI ·2026-11-14 ·Shiqi Huang, Weixi Yi, Shaheer Ullah Saeed, Yipei Wang, Wen Yan +7 more
similarity 0.69
Abstract

The objective of this study is to investigate the role of clinical variables - their breadth, source, and quality - in the development, utility, and generalisability of an AI system for bi-parametric MRI (bpMRI)-based diagnosis of clinically significant prostate cancer (csPCa). Clinical context is added to the published PI-CAI-2B image-only model, whose weights are fixed. The study cohort comprises 6,022 biparametric MRI cases from 4,540 patients at University College London Hospitals NHS Foundation Trust (2019-2024), of which 4,834 cases (3,632 patients) are held out for internal testing. Forty-three clinical variables are drawn from structured health records, clinical letters, MRI reports, and automated image measurements. The primary endpoint is the difference in the area under the receiver operating characteristic curve (AUROC) between the image-context and image-only models on the internal test set. Secondary analyses compare combining strategies, the contribution of individual variables, and performance across subgroups. All models are further tested on five external cohorts from primary diagnosis and population-based screening.

0.16

A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation

eess.IV cs.AIcs.CV ·2026-08-17 ·Siyuan Ma, Liang He, Mengying Zhu, Yi Chai, Mengyao Lyu +12 more
published similarity 0.67
Abstract

Missing or degraded sequences can limit prostate multiparametric MRI. We developed MSCNet, a sequence-conditioned cross-modal generative framework for reconstructing unavailable contrasts and restoring degraded acquisitions. Across ten completion tasks, task-specific MSCNet achieved mean structural similarity of 0.818 versus 0.798 for the strongest task-matched comparators; matched-capacity analyses showed larger differences in lesion fidelity and boundary preservation. In a blinded 1,000-case reader study, overall image quality met the prespecified non-inferiority criterion for DWI, ADC and T2W completion, but not T1W. In a separate 200-case diagnostic assessment, AUCs for clinically significant cancer were 0.860 with acquired images, 0.841 with MSCNet and 0.797 with baseline-generated images. A locked 186-case three-hospital cohort supported multicentre transportability. These retrospective results support quality-controlled cross-modal reconstruction as an adjunct to acquired prostate MRI.

0.47

TotalSynth: Robust Whole-Body Synthetic CT from MRI and CBCT

cs.CV ·2026-09-12 ·Valentin Boussot, Cedric Hemon, Anais Barateau, Caroline Lafond, Jean-Claude Nunes +1 more
similarity 0.64
Abstract · 10 pages, 3 figures

Purpose: To develop and evaluate TotalSynth, a reusable pretrained framework for whole-body synthetic CT (sCT) generation from MRI and cone-beam CT (CBCT) images. Materials and Methods: In this retrospective technical study, the dataset was assembled between 2020 and 2026 from SynthRAD challenge data, four prostate cohorts, and BIC-MAC. After registration quality control, 1450 of 1800 public challenge pairs were retained; 350 were excluded for insufficient registration quality or major source/CT mismatch. The corpus also included 84 additional prostate MRI/CT and 60 external BIC-MAC MRI/CT cases. Three 5-fold model families were evaluated with image-domain, anatomy-aware, registration-based, and uncertainty metrics. Age and sex were not consistently available across public datasets. Results: The released MRI-to-CT model achieved an overall MAE of 67.49 HU, SSIM of 0.920, and PSNR of 29.28 dB. The CBCT-to-CT model achieved an overall MAE of 53.55 HU, SSIM of 0.939, and PSNR of 32.09 dB. The unified model maintained similar performance on MRI inputs (MAE, 67.68 HU) and CBCT inputs (MAE, 54.22 HU). On external BIC-MAC data, MRI-to-CT MAE was 100.91 HU without fine-tuning and 62.21 HU after fine-tuning. Conclusion: TotalSynth provides reusable MRI- and CBCT-based CT synthesis models with broad anatomical coverage, while external evaluation highlights the need for local validation and optional fine-tuning under domain shift.

0.29

Distortion-Free High-Resolution PROPELLER-DWI: Technical Advances and Initial Demonstration for the Prostate

eess.IV ·2026-10-04 ·Jingjia Chen, Kun Zhou, Haoyang Pei, Angela Tong, Sherif Elsherif +5 more
similarity 0.64
Abstract · 37 pages, 10 figures

Purpose: To develop and evaluate an integrated acquisition and reconstruction framework for distortion-free, high-resolution prostate diffusion-weighted imaging (DWI). Methods: PROPELLER-DWI combines pulsed-gradient spin-echo diffusion preparation, an interleaved gradient-and-spin-echo PROPELLER acquisition with golden-angle blade rotation, and self-supervised joint reconstruction and denoising. It was compared with conventional NUFFT reconstruction and MPPCA denoising in a NIST/QIBA diffusion phantom, 25 volunteers, and 4 clinical patients undergoing 3T prostate MRI across three spatial-resolution settings and two b-values (50 and 1000 s/mm2). Reader studies compared geometric fidelity, image quality, and overall preference with clinical single-shot EPI-DWI. Results: PROPELLER-DWI preserved prostate geometry in cases where EPI-DWI showed susceptibility-induced distortion, signal loss, and metal-related artifacts. Self-supervised reconstruction improved noise suppression, with the greatest benefit at higher spatial resolution. Across all resolutions, PROPELLER-DWI had significantly higher geometric fidelity than EPI-DWI, and radiologist preference increased with the severity of EPI distortion. ADC measurements remained consistent across resolutions and comparable to clinical EPI-DWI, whereas EPI-DWI ADC varied with increasing resolution. Retrospective reduced-shot reconstruction also supported shorter scan times. Conclusion: PROPELLER-DWI enables distortion-free, high-resolution prostate DWI while preserving quantitative diffusion measurements, providing a practical alternative for technically challenging prostate MRI and applications requiring improved geometric fidelity.

0.24

A Data-Interventional Framework for Auditing Privacy and Fairness in Generative Medical Imaging

cs.CV ·2026-09-22 ·Mischa Dombrowski, Bernhard Kainz
top venuecode/data similarity 0.62
Abstract · Accepted for publication at the Journal of Machine Learning for Biomedical Imaging (MELBA) https://melba-journal.org/2026:026

Diffusion-based synthetic data generation offers a promising route for sharing medical imaging data without releasing sensitive patient records. However, generative models face a fundamental tension between privacy and fairness: they may memorize rare training samples, leading to privacy risks, or fail to reproduce underrepresented features, resulting in unfair synthetic distributions. While prior work has largely focused on either memorization or fairness in isolation, their interaction remains insufficiently understood. In this work, we introduce a data-interventional framework to systematically analyze privacy and fairness in diffusion models. We discuss synthetic anatomical fingerprints (SAFs), rare and manually injected image features, as controlled probes to study whether models generalize sensitive attributes across identities, memorize training samples, or suppress rare signals entirely. Across multiple conditioning modalities, we observe a consistent behavior: models either forget these fingerprints or memorize the entire image in which they appear, but do not generalize them to novel images. To support large-scale auditing where explicit sample extraction is infeasible, we further introduce the indicator metric t', which estimates a model's susceptibility to memorization by exploiting the internal structure of the diffusion process. By comparing conditioning signals of varying surprisal, we reveal a clear relationship between conditioning rarity and memorization behavior. Highly surprising conditioning signals act as retrieval keys that amplify memorization, whereas low-surprisal conditioning signals systematically suppress rare features, even when these appear repeatedly in the training data. Our findings provide actionable insights and concrete mitigation strategies for safe and fair synthetic medical data sharing. Code is available at https://github.com/MischaD/Privacy.

0.50

ProtoCouncil: Multi-Encoder Anatomical Prototypes for Prostate Cancer Detection in Micro-Ultrasound

cs.CV cs.AIcs.LG ·2026-11-06 ·Mohammad Mahdi Abootorabi, Nooshin Maghsoodi, Armin Saadat, Obed Dzikunu, Zhuoxin Guo +6 more
similarity 0.61
Abstract · 12 pages, 3 figures, Under Review

Purpose: Micro-ultrasound ($μ$US) is a promising modality for prostate cancer (PCa) detection, but identifying suspicious tissue depends heavily on clinical experience. Models trained on sparse, noisy core-level pathology offer little insight into their decisions and can fail under domain shift. Clinical use needs a traceable decision process and a flag for uncertain predictions. Methods: We introduce ProtoCouncil, in which three foundation encoders (MedSAM, DINOv3 and V-JEPA 2.1) learn case-based prototypes at three anatomical levels: whole gland, needle track and gland sectors. Prototypes are projected onto real training-core regions, and label-free fusion keeps every class logit as a bias plus an exact sum of prototype similarity contributions. Encoder disagreement triggers selective prediction. Results: In five-fold patient-level cross-validation (6,588 cores, 693 patients), ProtoCouncil reaches 82.9% AUROC for clinically significant PCa (csPCa, Grade Group $\geq$3), 4.3 points above the best previously published method. In zero-shot evaluation on two external cohorts, ProtoCouncil achieves higher all-cancer and csPCa AUROC than all previous methods. Nearest same-class exemplars match test-core tumor involvement better than random same-class cases. Externally, csPCa accuracy falls from 79% when encoders agree to 27% when two oppose the fused call. Abstaining on disagreement raises balanced accuracy by 7 and 15 points, while retaining 58% and 54% of the cores. Conclusion: ProtoCouncil grounds core-level predictions in exact similarity-based evidence from real, anatomically indexed training cases, which makes its reasoning interpretable. It achieves the highest csPCa AUROC across three cohorts and identifies unreliable predictions through encoder disagreement.

0.30