Share：
Share this content in WeChat
X
[Chinese] [PDF] 9 1
• Clinical Article •
The value of machine learning model constructed based on parameters of DCE-MRI and clinical risk factors in predicting the unstable state of rectal cancer microsatellites
TIAN Zhiwei  WANG Shilong  MAO Xijin 

DOI:10.12015/issn.1674-8034.2025.12.015.


[Abstract] Objective To explore the predictive value of a machine learning model integrating dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters and clinical risk factors for microsatellite instability (MSI) status in rectal cancer.Materials and Methods A retrospective analysis was conducted on 150 rectal cancer patients treated at Binzhou Medical University Hospital between May 2022 and July 2024, including 27 with MSI and 123 with microsatellite stability (MSS). MRI axial T2WI, apparent diffusion coefficient (ADC) maps, and dynamic contrast-enhanced T1WI (DCE-T1WI) were used to delineate the tumor's largest cross-sectional area as the region of interest (ROI). Radiomic features were extracted and reduced to identify optimal features. Independent clinical predictors and DCE parameters for MSI were selected using multivariate logistic regression. The dataset was split into training and validation sets in a 7∶3 ratio. Nine machine learning algorithms, extreme gradient boosting classifier (XGBoost), logistic regression (LR), light gradient boosting machine classifier (LGBM), random forest (RF), decision tree classifier (DT), Gaussian naive Bayes (GNB), support vector classifier (SVM), multilayer perceptron classifier (MLP), and adaptive boosting classifier (AdaBoost) were employed to construct predictive models. The performance of each model in predicting MSI status was evaluated using receiver operating characteristic (ROC) curves and decision curve analysis (DCA). Additionally, a temporal validation set comprising 30 rectal cancer patients from the same hospital between December 2024 and June 2025 was used to assess model generalizability via ROC analysis.Results The GNB model demonstrated the most stable performance. The combined model, incorporating independent clinical risk factors, radiomics scores, and DCE perfusion parameters, demonstrated superior predictive performance for MSI status in rectal cancer, with area under the curve (AUC) values of 0.920 (95% CI: 0.821 to 1.000), 0.900 (95% CI: 0.786 to 1.000), and 0.817 (95% CI: 0.667 to 0.966) in the training set, validation set, and temporal validation set, respectively.Conclusions Among the nine machine learning algorithms evaluated, GNB exhibited the best performance in predicting MSI status in rectal cancer, which was further validated using a temporal validation set. Machine learning models incorporating DCE-MRI parameters and clinical risk factors show promising value in predicting MSI status in rectal cancer.
[Keywords] rectal cancer；microsatellite instability；magnetic resonance imaging；dynamic contrast-enhanced；machine learning

TIAN Zhiwei1   WANG Shilong1   MAO Xijin1, 2*  

1 Department of Radiology, Binzhou Medical University Hospital, Binzhou 256600, China

2 School of Medical Imging, Binzhou Medical University, Yantai 264003, China

Corresponding author: MAO X J, E-mail: mao2483000@163.com

Conflicts of interest   None.

Received  2025-07-17
Accepted  2025-12-06
DOI: 10.12015/issn.1674-8034.2025.12.015
DOI:10.12015/issn.1674-8034.2025.12.015.

[1]
SIEGEL R L, MILLER K D, FUCHS H E, et al. Cancer statistics, 2022[J]. CA A Cancer J Clinicians, 2022, 72(1): 7-33. DOI: 10.3322/caac.21708.
[2]
ZHANG X, WU T, CAI X Y, et al. Neoadjuvant immunotherapy for MSI-H/dMMR locally advanced colorectal cancer: new strategies and unveiled opportunities[J/OL]. Front Immunol, 2022, 13: 795972 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/34307164/. DOI: 10.3389/fimmu.2022.795972.
[3]
MATSUBAYASHI H, OISHI T, SASAKI K, et al. Discordance of microsatellite instability and mismatch repair immunochemistry occurs depending on the cancer type[J]. Hum Pathol, 2023, 135: 54-64. DOI: 10.1016/j.humpath.2022.12.016.
[4]
BANDO H, TSUKADA Y, INAMORI K, et al. Preoperative chemoradiotherapy plus nivolumab before surgery in patients with microsatellite stable and microsatellite instability-high locally advanced rectal cancer[J]. Clin Cancer Res, 2022, 28(6): 1136-1146. DOI: 10.1158/1078-0432.CCR-21-3213.
[5]
TURNER K M, DELMAN A M, WIMA K, et al. Microsatellite instability is associated with worse overall survival in resectable colorectal liver metastases[J]. Am J Surg, 2023, 225(2): 322-327. DOI: 10.1016/j.amjsurg.2022.08.007.
[6]
LIU K, LUO R C, XU H J, et al. Comparative analysis of mismatch repair proteins and microsatellite instability detection in colorectal carcinoma[J]. Qingdao Med J, 2024, 56(3): 177-180. DOI: 10.3969/j.issn.1006-5571.2024.03.005.
[7]
GELSOMINO F, BARBOLINI M, SPALLANZANI A, et al. The evolving role of microsatellite instability in colorectal cancer: A review[J]. Cancer Treat Rev, 2016, 51: 19-26. DOI: 10.1016/j.ctrv.2016.10.005.
[8]
ZHANG W, HUANG Z X, ZHAO J, et al. Development and validation of magnetic resonance imaging-based radiomics models for preoperative prediction of microsatellite instability in rectal cancer[J/OL]. Ann Transl Med, 2021, 9(2): 134 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/33569436/. DOI: 10.21037/atm-20-7673.
[9]
SAMMAD A, DING Z X. Harnessing multi-omics: integrating radiomics and pathomics for predicting microsatellite instability in rectal cancer[J]. Acad Radiol, 2025, 32(4): 1946-1948. DOI: 10.1016/j.acra.2025.02.015.
[10]
CHEN W C, ZHENG K Y, YUAN W J, et al. A CT-based deep learning for segmenting tumors and predicting microsatellite instability in patients with colorectal cancers: a multicenter cohort study[J]. Radiol Med, 2025, 130(2): 214-225. DOI: 10.1007/s11547-024-01909-5.
[11]
YUAN H, PENG Y, XU X R, et al. A tumoral and peritumoral CT-based radiomics and machine learning approach to predict the microsatellite instability of rectal carcinoma[J]. Cancer Manag Res, 2022, 14: 2409-2418. DOI: 10.2147/CMAR.S377138.
[12]
ZHANG W, YIN H K, HUANG Z X, et al. Development and validation of MRI-based deep learning models for prediction of microsatellite instability in rectal cancer[J]. Cancer Med, 2021, 10(12): 4164-4173. DOI: 10.1002/cam4.3957.
[13]
LI Z B, DAI H, LIU Y X, et al. Radiomics analysis of multi-sequence MR images for predicting microsatellite instability status preoperatively in rectal cancer[J/OL]. Front Oncol, 2021, 11: 697497 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/34307164/. DOI: 10.3389/fonc.2021.697497.
[14]
CAI Z P, XU Z Y, CHEN Y F, et al. Multiparametric MRI subregion radiomics for preoperative assessment of high-risk subregions in microsatellite instability of rectal cancer patients: a multicenter study[J]. Int J Surg, 2024, 110(7): 4310-4319. DOI: 10.1097/JS9.0000000000001335.
[15]
BAI W Q, MA J F, LIU Y Y, et al. Screening of MSI detection loci and their heterogeneity in East Asian colorectal cancer patients[J]. Cancer Med, 2019, 8(5): 2157-2166. DOI: 10.1002/cam4.2111.
[16]
TAIEB J, SVRCEK M, COHEN R, et al. Deficient mismatch repair/microsatellite unstable colorectal cancer: Diagnosis, prognosis and treatment[J]. Eur J Cancer, 2022, 175: 136-157. DOI: 10.1016/j.ejca.2022.07.020.
[17]
CORNISH A J, GRUBER A J, KINNERSLEY B, et al. The genomic landscape of 2, 023 colorectal cancers[J]. Nature, 2024, 633(8028): 127-136. DOI: 10.1038/s41586-024-07747-9.
[18]
PENG L P, MA W T, ZHANG X L, et al. Predictive value of combined DCE-MRI perfusion parameters and clinical features nomogram for microsatellite instability in colorectal cancer[J/OL]. Discov Oncol, 2025, 16(1): 892 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/40410525/. DOI: 10.1007/s12672-025-02705-x.
[19]
YANG S Y, CHO M S, KIM N K. Difference between right-sided and left-sided colorectal cancers: from embryology to molecular subtype[J]. Expert Rev Anticancer Ther, 2018, 18(4): 351-358. DOI: 10.1080/14737140.2018.1442217.
[20]
MA Y Q, XU X R, LIN Y, et al. An integrative clinical and CT-based tumoral/peritumoral radiomics nomogram to predict the microsatellite instability in rectal carcinoma[J]. Abdom Radiol (NY), 2024, 49(3): 783-790. DOI: 10.1007/s00261-023-04099-2.
[21]
VANGALA D, NILIUS-ELILIWI V. Novel treatment concepts in patients with colorectal carcinomas and high microsatellite instability[J]. Zentralbl Chir, 2023, 148(6): 475-482. DOI: 10.1055/a-2012-4047.
[22]
CERCEK A, DOS SANTOS FERNANDES G, ROXBURGH C S, et al. Mismatch repair-deficient rectal cancer and resistance to neoadjuvant chemotherapy[J]. Clin Cancer Res, 2020, 26(13): 3271-3279. DOI: 10.1158/1078-0432.CCR-19-3728.
[23]
LI J J, XU Q, LUO C, et al. Clinicopathologic characteristics of resectable colorectal cancer with mismatch repair protein defects in Chinese population: Retrospective case series and literature review[J/OL]. Medicine (Baltimore), 2020, 99(24): e20554 [2025-08-10]. https://pubmed.ncbi.nlm.nih.gov/32541478/. DOI: 10.1097/MD.0000000000020554.
[24]
XING X W, LI D X, PENG J X, et al. A combinatorial MRI sequence-based radiomics model for preoperative prediction of microsatellite instability status in rectal cancer[J/OL]. Sci Rep, 2024, 14(1): 11760 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/38783014/. DOI: 10.1038/s41598-024-62584-0.
[25]
ZHANG Y, LIU J, WU C Y, et al. Preoperative prediction of microsatellite instability in rectal cancer using five machine learning algorithms based on multiparametric MRI radiomics[J/OL]. Diagnostics (Basel), 2023, 13(2): 269 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/36673079/. DOI: 10.3390/diagnostics13020269.
[26]
YAO X Z, DENG S T, HAN X Y, et al. Deep learning algorithm-based MRI radiomics and pathomics for predicting microsatellite instability status in rectal cancer: a multicenter study[J]. Acad Radiol, 2025, 32(4): 1934-1945. DOI: 10.1016/j.acra.2024.09.008.
[27]
WEI Z K, KANG Y J, PENG L P, et al. Prediction of microsatellite instability in colorectal cancer based on MRI-ADC and clinicopathological features[J]. Chin J Magn Reson Imag, 2025, 16(1): 48-53, 88. DOI: 10.12015/issn.1674-8034.2025.01.008.
[28]
LI Z, ZHANG J, ZHONG Q, et al. Development and external validation of a multiparametric MRI-based radiomics model for preoperative prediction of microsatellite instability status in rectal cancer: a retrospective multicenter study[J]. Eur Radiol, 2023, 33(3): 1835-1843. DOI: 10.1007/s00330-022-09160-0.
[29]
JING G D, CHEN Y K, MA X L, et al. Predicting mismatch-repair status in rectal cancer using multiparametric MRI-based radiomics models: A preliminary study[J/OL]. Biomed Res Int, 2022, 2022: 6623574 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/27009765/. DOI: 10.1155/2022/6623574.
[30]
TOH J, CHAPUIS P H, BOKEY L, et al. Competing risks analysis of microsatellite instability as a prognostic factor in colorectal cancer[J]. Br J Surg, 2017, 104(9): 1250-1259. DOI: 10.1002/bjs.10542.
[31]
KIM H R, KIM S H, NAM K H. Association between dynamic contrast-enhanced MRI parameters and prognostic factors in patients with primary rectal cancer[J]. Curr Oncol, 2023, 30(2): 2543-2554. DOI: 10.3390/curroncol30020194.
[32]
AO W Q, ZHANG X, YAO X Z, et al. Preoperative prediction of extramural venous invasion in rectal cancer by dynamic contrast-enhanced and diffusion weighted MRI: a preliminary study[J/OL]. BMC Med Imaging, 2022, 22(1): 78 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/35484509/. DOI: 10.1186/s12880-022-00810-9.
[33]
SHI L Y, PAN N N, ZHAO J X, et al. Study on the prediction of microsatellite instability in colorectal cancer based on DCE MRI and clinical features[J]. Chin Comput Med Imag, 2024, 30(4): 473-484. DOI: 10.19627/j.cnki.cn31-1700/th.2024.04.013.
[34]
ZHAO B S, TAN Y Q, TSAI W Y, et al. Reproducibility of radiomics for deciphering tumor phenotype with imaging[J]. Sci Rep, 2016, 6: 23428. DOI: 10.1038/srep23428.
[35]
WYNANTS L, VAN CALSTER B, COLLINS G S, et al. Prediction models for diagnosis and prognosis of covid-19: systematic review and critical appraisal[J/OL]. BMJ, 2020, 369: m1328 [2025-05-20]. https://pubmed.ncbi.nlm.nih.gov/32265220/. DOI: 10.1136/bmj.m1328.

PREV The diagnostic value of 5.0 T ultra-high-field MRI susceptibility weighted imaging combined with multi-echo T2<sup>*</sup> mapping in microvascular invasion of hepatocellular carcinoma
NEXT The value of nomogram model based on IVIM-MRI radiomics for the noninvasive assessment of renal fibrosis in chronic kidney disease
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn