---
title: "MSCNet: Cross‑modal generative reconstruction for incomplete and degraded prostate MRI"
id: "biorxiv-7-a-cross-modal-generative-model-for-incomplete-and-degradedprostate-mri-with"
canonical_url: "https://medichelpline.com/clinical-feed/biorxiv-7-a-cross-modal-generative-model-for-incomplete-and-degradedprostate-mri-with"
content_type: "clinical_feed_article"
specialty: "Radiology"
source_name: "bioRxiv (Biomedical Preprints)"
source_url: "https://www.biorxiv.org/content/10.64898/2026.08.16.745066v1?rss=1"
published_at: "2026-08-18T12:00:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# MSCNet: Cross‑modal generative reconstruction for incomplete and degraded prostate MRI
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/biorxiv-7-a-cross-modal-generative-model-for-incomplete-and-degradedprostate-mri-with
- **Specialty:** [Radiology](https://medichelpline.com/clinical-feed/radiology.md)
- **Primary Source:** bioRxiv (Biomedical Preprints)
- **Source URL:** [Original Journal Publication](https://www.biorxiv.org/content/10.64898/2026.08.16.745066v1?rss=1)
- **Published At:** 2026-08-18T12:00:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- Missing or degraded sequences limit the utility of multiparametric **prostate MRI**; the authors developed a sequence-conditioned cross‑modal generative framework named **MSCNet** to reconstruct unavailable contrasts and restore degraded acquisitions. - MSCNet was evaluated across ten distinct completion tasks; task‑specific versions achieved mean structural similarity (SSIM) of **0.818**, compared with **0.798** for the strongest task‑matched comparator. - Matched‑capacity analyses indicated larger improvements with MSCNet in **lesion fidelity** and **boundary preservation** relative to comparators. - A blinded 1,000‑case reader study tested image quality: MSCNet completion met the prespecified non‑inferiority criterion for **DWI**, **ADC**, and **T2W** completion, but did not meet non‑inferiority for **T1W** completion. - In a separate 200‑case diagnostic assessment of clinically significant prostate cancer, AUCs were reported as **0.860** with acquired images, **0.841** with MSCNet‑generated images, and **0.797** with baseline‑generated images. - A locked 186‑case cohort spanning three hospitals supported **multicentre transportability** of the approach. - The study reports retrospective results suggesting that **quality‑controlled cross‑modal reconstruction** can be used as an adjunct to acquired prostate MRI when sequences are missing or degraded. - The authors declared no competing interests. Specific implementation details, training data composition, exact architecture parameters, and prospective clinical impact beyond the reported retrospective cohorts were not reported in the source abstract.
## Clinical Analysis & Structured Key Points
A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation | bioRxiv Skip to main content New Results A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation SiYuan Ma , Liang He , Mengying Zhu , Yi Chai , Mengyao Lyu , Haowei Wang , Qizhen Lan , HaoBo Sun , Qixin Zhang , Jingli Chen , Xiaobing Wei , Jiaming Liu , Guiqin Liu , Qianwen Zhang , Yang Liu , Dacheng Tao , Guangyu Wu doi: https://doi.org/10.64898/2026.08.16.745066 SiYuan Ma 1 Nanyang Technological University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Liang He 2 Stanford University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mengying Zhu 3 Shanghai Jiao Tong University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yi Chai 4 National University of Singapore; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mengyao Lyu 1 Nanyang Technological University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Haowei Wang 5 Tongji University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Qizhen Lan 6 The University of Texas Health Science Center at Houston (UTHealth Houston); Find this author on Google Scholar Find this author on PubMed Search for this author on this site HaoBo Sun 7 Southeast University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Qixin Zhang 1 Nanyang Technological University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jingli Chen 3 Shanghai Jiao Tong University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Xiaobing Wei 3 Shanghai Jiao Tong University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jiaming Liu 8 Ningbo Hangzhou Bay Hospital; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Guiqin Liu 3 Shanghai Jiao Tong University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Qianwen Zhang 9 Naval Medical University Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yang Liu 1 Nanyang Technological University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Dacheng Tao 1 Nanyang Technological University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: dacheng.tao{at}ntu.edu.sg Guangyu Wu 3 Shanghai Jiao Tong University School of Medicine; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abstract Info/History Metrics Preview PDF 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. Competing Interest Statement The authors have declared no competing interest. Copyright The copyright holder for this preprint is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY 4.0 International license . Back to top Previous Posted August 18, 2026. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation SiYuan Ma , Liang He , Mengying Zhu , Yi Chai , Mengyao Lyu , Haowei Wang , Qizhen Lan , HaoBo Sun , Qixin Zhang , Jingli Chen , Xiaobing Wei , Jiaming Liu , Guiqin Liu , Qianwen Zhang , Yang Liu , Dacheng Tao , Guangyu Wu bioRxiv 2026.08.16.745066; doi: https://doi.org/10.64898/2026.08.16.745066 Share This Article: Copy Citation Tools A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation SiYuan Ma , Liang He , Mengying Zhu , Yi Chai , Mengyao Lyu , Haowei Wang , Qizhen Lan , HaoBo Sun , Qixin Zhang , Jingli Chen , Xiaobing Wei , Jiaming Liu , Guiqin Liu , Qianwen Zhang , Yang Liu , Dacheng Tao , Guangyu Wu bioRxiv 2026.08.16.745066; doi: https://doi.org/10.64898/2026.08.16.745066 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Areas All Articles Animal Behavior and Cognition (7899) Biochemistry (18485) Bioengineering (14643) Bioinformatics (43723) Biophysics (22250) Cancer Biology (19383) Cell Biology (26523) Clinical Trials (138) Developmental Biology (13802) Ecology (20685) Epidemiology (2067) Evolutionary Biology (25130) Genetics (16003) Genomics (23250) Immunology (18430) Microbiology (41901) Molecular Biology (17796) Neuroscience (92082) Paleontology (688) Pathology (2940) Pharmacology and Toxicology (5021) Physiology (7993) Plant Biology (15752) Scientific Communication and Education (2085) Synthetic Biology (4495) Systems Biology (10114) Zoology (2354)
## Related Clinical Research

- [NICE recommends five AI tools to support prostate cancer detection on MRI](https://medichelpline.com/clinical-feed/bmj-0-five-ai-tools-for-prostate-cancer-screening-are-recommended-by-nice.md)
- [Micro-Ultrasound Foundation Model (muFM) Enables Data-Efficient Prostate Image Analysis](https://medichelpline.com/clinical-feed/medrxiv-9-data-efficient-prostate-micro-ultrasound-image-analysis-using-a-modality.md)
- [Gated-Attention Multiple Instance Learning for Multi-Center Cervical Cytology Slide Triage](https://medichelpline.com/clinical-feed/medrxiv-1-weakly-supervised-gated-attention-multiple-instance-learning-for-multi-center.md)
- [Formononetin targets GSTM2 to reshape the prostate cancer immune microenvironment via ROS/STAT3 ch](https://medichelpline.com/clinical-feed/frontiers-in-immunology-9-integrated-pharmacology-and-validation-reveal-formononetin-upregulates-gstm2-to.md)
- [Editorial Expression of Concern: Image concerns in EphA2–arachidonic acid prostate cancer paper](https://medichelpline.com/clinical-feed/british-journal-of-cancer-0-editorial-expression-of-concern-ligand-independent-activation-of-epha2-by.md)

## Navigation
- [← Back to Radiology Feed](https://medichelpline.com/clinical-feed/radiology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.