Share：
Share this content in WeChat
X
[Chinese] [PDF] 10 1
• Clinical Article •
Prediction of pathological grading in prostate adenocarcinoma based on multiparametric MRI habitat imaging
ZHAI Chengfeng  YANG Hongkai  QI Xuan  YANG Xin  QI Dong  CHENG Weiqun  HE Yongsheng 

DOI:10.12015/issn.1674-8034.2025.12.017.


[Abstract] Objective To investigate the value of biparametric magnetic resonance imaging (bpMRI) habitat imaging analysis in diagnosing high-risk and low-risk prostate cancner (PCa).Materials and Methods A retrospective analysis was conducted on 191 PCa patients confirmed by biopsy or surgical pathology at Ma'anshan People's Hospital between December 2023 and August 2024, including 131 high-risk PCa cases and 60 low-risk PCa cases. The 191 patients were randomly divided into training and testing sets at a 7∶3 ratio. All patients underwent bpMRI scans, with preprocessing performed on T2WI, ZOOMit diffusion weighted imaging (ZOOMit-DWI), and apparent diffusion coefficient (ADC) sequences. Nineteen radiomics features were extracted from ADC images. By integrating T2WI and ZOOMit-DWI images, unsupervised K-means clustering was used to generate similar subregions across all tumor voxels. Based on the habitat subregion results, the intratumoral heterogeneity score (ITHscore) was calculated for the 191 patients. Radiomics features were extracted from the subregions, followed by dimensionality reduction and filtering to select features with the highest correlation coefficients. SHAP analysis was employed to visualize feature importance. After feature fusion and selection, 10 habitat radiomics models were established. For each model, the threshold, sensitivity, specificity, accuracy, negative predictive value, and positive predictive value were calculated. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was computed to evaluate diagnostic performance. Decision curve analysis (DCA) was further used to assess the net benefit of the models.Results In both training and testing sets, the difference in total prostate specific antigen (tPSA) between high-risk and low-risk PCa was statistically significant. Based on the Calinski-Harabasz (CH) value, 2 was determined as the optimal number of habitat subregions. SHAP analysis revealed that the original_glszm_ZoneEntropy feature in the h2 subregion had the greatest impact. A total of 10 classifiers were employed for model construction. The habitat radiomics model was compared with the clinical + prostate imaging reporting and data system (PI-RADS) score model. In the habitat radiomics model, the extremely randomized trees (ExtraTrees) model demonstrated the best prediction performance in the test set, with a training set AUC of 0.838 [95% (confidence interval, CI): 0.768 to 0.908] and a test set AUC of 0.796 (95% CI: 0.665 to 0.927). For the model constructed using clinical data and PI-RADS scores, the logistic regression (LR) model achieved the highest predictive performance, with a training set AUC of 0.786 (95% CI: 0.705 to 0.866) and a test set AUC of 0.719 (95% CI: 0.550 to 0.887). The clinical net benefit of both models was evaluated using DCA, and the habitat radiomics model exhibited both a higher AUC value range than the clinical-PI-RADS score model.Conclusions Based on bpMRI habitat imaging analysis, the pathological classification of PCa can be diagnosed relatively accurately, which is helpful for the clinical diagnosis and risk prediction of PCa.
[Keywords] prostate cancer；habitat imaging；intra-tumor heterogeneity；cluster analysis；magnetic resonance imaging

ZHAI Chengfeng1, 2, 3   YANG Hongkai2, 3   QI Xuan2, 3   YANG Xin1, 2, 3   QI Dong1, 2, 3   CHENG Weiqun1, 2, 3   HE Yongsheng1, 2, 3*  

1 Department of Radiology, Maanshan People's Hospital, Anhui Medical University, Maanshan 243000, China

2 Department of Radiology, Maanshan People's Hospital affiliated to Wannan Medical College, Maanshan 243000, China

3 Maanshan Key Laboratory for Medical Image Modeling and Intelligent Analysis, Maanshan 243099, China

Corresponding author: HE Y S, E-mail: heyongsheng881@163.com

Conflicts of interest   None.

Received  2025-08-03
Accepted  2025-12-06
DOI: 10.12015/issn.1674-8034.2025.12.017
DOI:10.12015/issn.1674-8034.2025.12.017.

[1]
JAMES N D, TANNOCK I, N'DOW J, et al. The lancet commission on prostate cancer: planning for the surge in cases[J]. Lancet, 2024, 403(10437): 1683-1722. DOI: 10.1016/S0140-6736(24)00651-2.
[2]
ZHENG R S, CHEN R, HAN B F, et al. Cancer incidence and mortality in China, 2022[J]. Chin J Oncol, 2024, 46(3): 221-231. DOI: 10.3760/cma.j.cn112152-20240119-00035.
[3]
KAMEL M H, KHALIL M I, ALOBUIA W M, et al. Incidence of metastasis and prostate-specific antigen levels at diagnosis in Gleason 3+4 versus 4+3 prostate cancer[J]. Urol Ann, 2018, 10(2): 203-208. DOI: 10.4103/UA.UA_124_17.
[4]
LAUNER B M, ELLIS T A, SCARPATO K R. A contemporary review: mpMRI in prostate cancer screening and diagnosis[J]. Urol Oncol, 2025, 43(1): 15-22. DOI: 10.1016/j.urolonc.2024.05.012.
[5]
PAN Y S, SHEN C, CHEN X F, et al. bpMRI and mpMRI for detecting prostate cancer: A retrospective cohort study[J/OL]. Front Surg, 2023, 9: 1096387 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/36726941/. DOI: 10.3389/fsurg.2022.1096387.
[6]
SHI Z W, LIU Z Y. Predicament and outlet of imageology research[J]. Chin J Radiol, 2022, 56(1): 9-11. DOI: 10.3760/cma.j.cn112149-20211111-00998.
[7]
CHENG W Q, QI X, YANG H K, et al. Identification of type Luminal and non-type Luminal breast cancers based on multiparametric MR habitat imaging[J]. Chin J Magn Reson Imag, 2025, 16(5): 170-180. DOI: 10.12015/issn.1674-8034.2025.05.026.
[8]
DEL N J FLORES-TÉLLEZ T, BAENA E. Experimental challenges to modeling prostate cancer heterogeneity[J]. Cancer Lett, 2022, 524: 194-205. DOI: 10.1016/j.canlet.2021.10.012.
[9]
GILLIES R J, BALAGURUNATHAN Y. Perfusion MR imaging of breast cancer: insights using "habitat imaging"[J]. Radiology, 2018, 288(1): 36-37. DOI: 10.1148/radiol.2018180271.
[10]
YANG Z T, WU H, GAO H Y, et al. Progress in the application of habitat imaging in multi-system tumors[J]. Chin J Magn Reson Imag, 2025, 16(3): 222-227. DOI: 10.12015/issn.1674-8034.2025.03.038.
[11]
ZHAO B, JU S H. Habitat imaging: exploring tumor heterogeneity[J]. Chin J Radiol, 2025, 59(4): 355-357. DOI: 10.3760/cma.j.cn112149-20241230-00767.
[12]
YIN X Y, SHA H, CAO X J, et al. Tumor habitat-derived radiomics features in pretreatment CT scans for predicting concurrent chemoradiotherapy responses in nasopharyngeal carcinoma: a retrospective study[J]. Quant Imaging Med Surg, 2025, 15(4): 2917-2928. DOI: 10.21037/qims-24-1642.
[13]
LI J Q, QIU Z B, ZHANG C, et al. ITHscore: comprehensive quantification of intra-tumor heterogeneity in NSCLC by multi-scale radiomic features[J]. Eur Radiol, 2023, 33(2): 893-903. DOI: 10.1007/s00330-022-09055-0.
[14]
KANE C J, EGGENER S E, SHINDEL A W, et al. Variability in outcomes for patients with intermediate-risk prostate cancer (gleason score 7, international society of urological pathology gleason group 2-3) and implications for risk stratification: a systematic review[J]. Eur Urol Focus, 2017, 3(4/5): 487-497. DOI: 10.1016/j.euf.2016.10.010.
[15]
STEPHAN C, RITTENHOUSE H, HU X H, et al. Prostate-specific antigen (PSA) screening and new biomarkers for prostate cancer (PCa)[J]. EJIFCC, 2014, 25(1): 55-78.
[16]
MAHMOUD M M, ABDEL HAMID F F, ABDELGAWAD I, et al. Diagnostic efficacy of PSMA and PSCA mRNAs combined to PSA in prostate cancer patients[J]. Asian Pac J Cancer Prev, 2023, 24(1): 223-229. DOI: 10.31557/APJCP.2023.24.1.223.
[17]
FERNANDES M C, YILDIRIM O, WOO S, et al. The role of MRI in prostate cancer: current and future directions[J]. MAGMA, 2022, 35(4): 503-521. DOI: 10.1007/s10334-022-01006-6.
[18]
O'SHEA A, HARISINGHANI M. PI-RADS: multiparametric MRI in prostate cancer[J]. MAGMA, 2022, 35(4): 523-532. DOI: 10.1007/s10334-022-01019-1.
[19]
IACOB R, MANOLESCU D, STOICESCU E R, et al. The diagnostic value of bpMRI in prostate cancer: benefits and limitations compared to mpMRI[J/OL]. Bioengineering (Basel), 2024, 11(10): 1006 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/39451382/. DOI: 10.3390/bioengineering11101006.
[20]
BREMBILLA G, GIGANTI F, SIDHU H, et al. Diagnostic accuracy of abbreviated bi-parametric MRI (a-bpMRI) for prostate cancer detection and screening: A multi-reader study[J/OL]. Diagnostics (Basel), 2022, 12(2): 231 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/35204322/. DOI: 10.3390/diagnostics12020231.
[21]
TAMADA T, UEDA Y, UENO Y, et al. Diffusion-weighted imaging in prostate cancer[J]. MAGMA, 2022, 35(4): 533-547. DOI: 10.1007/s10334-021-00957-6.
[22]
TEICĂ R V, ȘERBĂNESCU M S, FLORESCU L M, et al. Tumor area highlighting using T2WI, ADC map, and DWI sequence fusion on bpMRI images for better prostate cancer diagnosis[J/OL]. Life (Basel), 2023, 13(4): 910 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/37109440/. DOI: 10.3390/life13040910.
[23]
LIU X, WANG X P, ZHANG Y F, et al. Preoperative prediction of pelvic lymph nodes metastasis in prostate cancer using an ADC-based radiomics model: comparison with clinical nomograms and PI-RADS assessment[J]. Abdom Radiol (NY), 2022, 47(9): 3327-3337. DOI: 10.1007/s00261-022-03583-5.
[24]
LIU Y, WANG W, QIN X B, et al. The applied research of simultaneous image acquisition of T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) in the assessment of patients with prostate cancer[J]. Asian J Androl, 2019, 21(2): 177-182. DOI: 10.4103/aja.aja_82_18.
[25]
KRISHNA S, SCHIEDA N, MCINNES M D, et al. Diagnosis of transition zone prostate cancer using T2-weighted (T2W) MRI: comparison of subjective features and quantitative shape analysis[J]. Eur Radiol, 2019, 29(3): 1133-1143. DOI: 10.1007/s00330-018-5664-z.
[26]
SHAISH H, KANG S K, ROSENKRANTZ A B. The utility of quantitative ADC values for differentiating high-risk from low-risk prostate cancer: a systematic review and meta-analysis[J]. Abdom Radiol (NY), 2017, 42(1): 260-270. DOI: 10.1007/s00261-016-0848-y.
[27]
WOO S, KIM S Y, CHO J Y, et al. Preoperative evaluation of prostate cancer aggressiveness: using ADC and ADC ratio in determining gleason score[J]. AJR Am J Roentgenol, 2016, 207(1): 114-120. DOI: 10.2214/AJR.15.15894.
[28]
HUANG Y, WANG X X, CAO Y, et al. Nomogram for predicting neoadjuvant chemotherapy response in breast cancer using MRI-based intratumoral heterogeneity quantification[J/OL]. Radiology, 2025, 315(1): e241805 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/40232145/. DOI: 10.1148/radiol.241805.
[29]
GENG T, WU Z Q, QIN G G, et al. Research progress of radiomics and radiogenomics in predicting the molecular subtypes of breast cancer[J]. Int J Med Radiol, 2024, 47(3): 306-309. DOI: 10.19300/j.2024.Z21290.
[30]
LIU Y H, GAO Y. Implications of habitat imaging-based multisequence MRI in adult-type diffuse glioma[J]. Chin J Magn Reson Imag, 2023, 14(9): 119-124. DOI: 10.12015/issn.1674-8034.2023.09.022.
[31]
FINGER A M, HENDLEY A M, FIGUEROA D, et al. Tissue mechanics in tumor heterogeneity and aggression[J]. Trends Cancer, 2025, 11(8): 806-824. DOI: 10.1016/j.trecan.2025.04.004.
[32]
LI S L, DAI Y M, CHEN J Y, et al. MRI-based habitat imaging in cancer treatment: current technology, applications, and challenges[J/OL]. Cancer Imaging, 2024, 24(1): 107 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/39148139/. DOI: 10.1186/s40644-024-00758-9.
[33]
BAI R L, CUI J W. Tumor heterogeneity: the great challenge in precision clinical diagnosis and treatment[J]. Chin J Clin Oncol, 2020, 47(21): 1082-1087. DOI: 10.3969/j.issn.1000-8179.2020.21.341.
[34]
WU H J, TEMKO D, MALIGA Z, et al. Spatial intra-tumor heterogeneity is associated with survival of lung adenocarcinoma patients[J/OL]. Cell Genom, 2022, 2(8): 100165 [2025-01-25]. https://pubmed.ncbi.nlm.nih.gov/36419822/. DOI: 10.1016/j.xgen.2022.100165.
[35]
YANG Q F, LIU Y Y, ZENG Y L, et al. Mechanism, detection and clinical implication of tumor heterogeneity[J]. Int J Med Radiol, 2017, 44(12): 922-925. DOI: 10.3760/cma.j.issn.1673-422X.2017.12.010.
[36]
SHAH R B, ZHOU M. Recent advances in prostate cancer pathology: Gleason grading and beyond[J]. Pathol Int, 2016, 66(5): 260-272. DOI: 10.1111/pin.12398.
[37]
KANG Z, XU A H, WANG L. Predictive role of T2WI and ADC-derived texture parameters in differentiating Gleason score 3 + 4 and 4 + 3 prostate cancer[J]. J Xray Sci Technol, 2021, 29(2): 307-315. DOI: 10.3233/XST-200785.
[38]
YADAV S S, STOCKERT J A, HACKERT V, et al. Intratumor heterogeneity in prostate cancer[J]. Urol Oncol, 2018, 36(8): 349-360. DOI: 10.1016/j.urolonc.2018.05.008.
[39]
PARRA N A, LU H, CHOI J, et al. Habitats in DCE-MRI to predict clinically significant prostate cancers[J]. Tomography, 2019, 5(1): 68-76. DOI: 10.18383/j.tom.2018.00037.
[40]
YUAN L, ZHANG J L, MA L N, et al. Non-invasive quantitative visualization of multi-parametric MRI habitat imaging for predicting prostate cancer risk degree[J]. Chin J Radiol, 2025, 59(4): 393-400. DOI: 10.3760/cma.j.cn112149-20240624-00348.

PREV The value of nomogram model based on IVIM-MRI radiomics for the noninvasive assessment of renal fibrosis in chronic kidney disease
NEXT Diagnostic value of synthetic magnetic resonance imaging combined with amide proton transfer-weighted imaging in the grading of prostate cancer
  


LINK


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