MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI

Accurate preoperative prediction of biochemical recurrence (BCR) in prostate cancer (PCa) is essential for treatment optimization, and demands an explicit focus on tumor microenvironment (TME). To address this, we developed MRMS-CNNFormer, an innovative framework integrating 2D <b>m</b>u...

Full description

Saved in:
Bibliographic Details
Main Authors: Tao Lian, Mengting Zhou, Yangyang Shao, Xiaqing Chen, Yinghua Zhao, Qianjin Feng
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Bioengineering
Subjects:
Online Access:https://www.mdpi.com/2306-5354/12/5/538
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849327455905513472
author Tao Lian
Mengting Zhou
Yangyang Shao
Xiaqing Chen
Yinghua Zhao
Qianjin Feng
author_facet Tao Lian
Mengting Zhou
Yangyang Shao
Xiaqing Chen
Yinghua Zhao
Qianjin Feng
author_sort Tao Lian
collection DOAJ
description Accurate preoperative prediction of biochemical recurrence (BCR) in prostate cancer (PCa) is essential for treatment optimization, and demands an explicit focus on tumor microenvironment (TME). To address this, we developed MRMS-CNNFormer, an innovative framework integrating 2D <b>m</b>ulti-<b>r</b>egion (intratumoral, peritumoral, and periprostatic) and <b>m</b>ulti-<b>s</b>equence magnetic resonance imaging (MRI) images (T2-weighted imaging with fat suppression (T2WI-FS) and diffusion-weighted imaging (DWI)) with clinical characteristics. The framework utilizes a <b>CNN</b>-based encoder for imaging feature extraction, followed by a trans<b>former</b>-based encoder for multi-modal feature integration, and ultimately employs a fully connected (FC) layer for final BCR prediction. In this multi-center study (46 BCR-positive cases, 186 BCR-negative cases), patients from centers A and B were allocated to training (<i>n</i> = 146) and validation (<i>n</i> = 36) sets, while center C patients (<i>n</i> = 50) formed the external test set. The multi-region MRI-based model demonstrated superior performance (AUC, 0.825; 95% CI, 0.808–0.852) compared to single-region models. The integration of clinical data further enhanced the model’s predictive capability (AUC 0.835; 95% CI, 0.818–0.869), significantly outperforming the clinical model alone (AUC 0.612; 95% CI, 0.574–0.646). MRMS-CNNFormer provides a robust, non-invasive approach for BCR prediction, offering valuable insights for personalized treatment planning and clinical decision making in PCa management.
format Article
id doaj-art-08b67a33a250481db05b39fb42340367
institution Kabale University
issn 2306-5354
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Bioengineering
spelling doaj-art-08b67a33a250481db05b39fb423403672025-08-20T03:47:53ZengMDPI AGBioengineering2306-53542025-05-0112553810.3390/bioengineering12050538MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRITao Lian0Mengting Zhou1Yangyang Shao2Xiaqing Chen3Yinghua Zhao4Qianjin Feng5School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, ChinaDepartment of Radiology, The Third Affiliated Hospital, Southern Medical University, Guangzhou 510630, ChinaDepartment of Radiology, The Third Affiliated Hospital, Southern Medical University, Guangzhou 510630, ChinaDepartment of Radiology, The Second Clinical Medical College of Guangzhou University of Chinese Medicine, Guangdong Provincial Hospital of Traditional Chinese Medicine, Guangzhou 510210, ChinaDepartment of Radiology, The Third Affiliated Hospital, Southern Medical University, Guangzhou 510630, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou 510515, ChinaAccurate preoperative prediction of biochemical recurrence (BCR) in prostate cancer (PCa) is essential for treatment optimization, and demands an explicit focus on tumor microenvironment (TME). To address this, we developed MRMS-CNNFormer, an innovative framework integrating 2D <b>m</b>ulti-<b>r</b>egion (intratumoral, peritumoral, and periprostatic) and <b>m</b>ulti-<b>s</b>equence magnetic resonance imaging (MRI) images (T2-weighted imaging with fat suppression (T2WI-FS) and diffusion-weighted imaging (DWI)) with clinical characteristics. The framework utilizes a <b>CNN</b>-based encoder for imaging feature extraction, followed by a trans<b>former</b>-based encoder for multi-modal feature integration, and ultimately employs a fully connected (FC) layer for final BCR prediction. In this multi-center study (46 BCR-positive cases, 186 BCR-negative cases), patients from centers A and B were allocated to training (<i>n</i> = 146) and validation (<i>n</i> = 36) sets, while center C patients (<i>n</i> = 50) formed the external test set. The multi-region MRI-based model demonstrated superior performance (AUC, 0.825; 95% CI, 0.808–0.852) compared to single-region models. The integration of clinical data further enhanced the model’s predictive capability (AUC 0.835; 95% CI, 0.818–0.869), significantly outperforming the clinical model alone (AUC 0.612; 95% CI, 0.574–0.646). MRMS-CNNFormer provides a robust, non-invasive approach for BCR prediction, offering valuable insights for personalized treatment planning and clinical decision making in PCa management.https://www.mdpi.com/2306-5354/12/5/538prostate cancerbiochemical recurrencemagnetic resonance imagingdeep learningpreoperative prognostic prediction
spellingShingle Tao Lian
Mengting Zhou
Yangyang Shao
Xiaqing Chen
Yinghua Zhao
Qianjin Feng
MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
Bioengineering
prostate cancer
biochemical recurrence
magnetic resonance imaging
deep learning
preoperative prognostic prediction
title MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
title_full MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
title_fullStr MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
title_full_unstemmed MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
title_short MRMS-CNNFormer: A Novel Framework for Predicting the Biochemical Recurrence of Prostate Cancer on Multi-Sequence MRI
title_sort mrms cnnformer a novel framework for predicting the biochemical recurrence of prostate cancer on multi sequence mri
topic prostate cancer
biochemical recurrence
magnetic resonance imaging
deep learning
preoperative prognostic prediction
url https://www.mdpi.com/2306-5354/12/5/538
work_keys_str_mv AT taolian mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri
AT mengtingzhou mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri
AT yangyangshao mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri
AT xiaqingchen mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri
AT yinghuazhao mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri
AT qianjinfeng mrmscnnformeranovelframeworkforpredictingthebiochemicalrecurrenceofprostatecanceronmultisequencemri