Constructing a Prediction Model for Clinically Significant Prostate Cancer Combined with Radiomics Features of MRI and PRKY Promoter Methylation Level in Urine Samples

Constructing a Prediction Model for Clinically Significant Prostate Cancer Combined with Radiomics Features of MRI and PRKY Promoter Methylation Level in Urine Samples | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Constructing a Prediction Model for Clinically Significant Prostate Cancer Combined with Radiomics Features of MRI and PRKY Promoter Methylation Level in Urine Samples Zheng Dai, Yufei Wang, Zeyu Chen, Guosheng Xu, Song Sun, Wei Liu, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6806278/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Purpose To enhance the diagnostic value of MRI for clinically significant prostate cancer (csPCa) and optimize the diagnostic process of prostate cancer (PCa), we developed a machine learning-based prediction model for csPCa combined with MRI features, clinical data, and PRKY promoter methylation level in urine samples. Methods Thirty-nine patients who underwent prostate biopsy or transurethral laser enucleation of the prostate from 2022 to 2023 were selected for this study, and their clinical data and MRI images were obtained before the operation. The urine samples of these patients were collected after prostate massage. Methylation level of two PRKY promoter sites, cg05618150 and cg05163709, were tested through Methylation-Specific PCR. The PI-RADS score of each patient was estimated and the region of interest (ROI) was delineated. After being extracted by a plug-in of 3D-slicer, radiomics features were selected through LASSO regression and t-test. Selected radiomic features, methylation levels and clinical data were used for model construction through the random forest (RF) algorithm in Python. The model based on the PI-RADS score was also constructed for comparison with the radiomics model. The predictive efficiency of each model was analyzed by the area under the receiver operation characteristic (ROC) curve (AUC), and all the models have gone through 3-fold cross-validation. Results Methylation level of cg05163709 in csPCa patients was higher than that in clinically insignificant PCa and benign prostatic hyperplasia patients. The AUC of cg05163709 in csPCA prediction was 0.75. The AUC of the model combined with T2WI and ADC features was 0.91. And the model combined with radiomics features, Methylation level of cg05163709 and clinical data reached an AUC of 0.97, which was greater than that of the model based on the PI-RADS score (AUC = 0.86). Conclusion An effective prediction model for csPCa was successfully established by integrating T2WI features, clinical data, and methylation level of cg05163709 in urine specimens. Prostate cancer Radiomics PRKY DNA methylation Urine specimen Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Prostate cancer (PCa) is one of the most common cancer in males and the leading cause of cancer-related mortality among men in many countries( 1 ). Due to its insidious symptoms, many patients are diagnosed at advanced stages. Thus, early and accurate diagnosis of PCa is crucial. Pathological examination of prostate tissue is the gold standard for PCa diagnosis( 2 ). According to the International Society of Urological Pathology (ISUP) grading system, PCa is classified into grades 1 to 5 based on tumor features. The 2024 European Association of Urology (EAU) guidelines defined ISUP grade 1 (Gleason score 3 + 3 = 6) as clinically insignificant prostate cancer (cisPCa), while tumors with ISUP grades ≥ 2 were classified as clinically significant prostate cancer (csPCa)( 2 ). These two categories exhibit differences in prognosis. The EAU guidelines suggest that cisPCa showes better prognosis than csPCa, active surveillance should be considered firstly rather than early intervention. In contrast, csPCa patients require prompt surgery or endocrine therapy once diagnosed( 2 – 4 ). Diagnosis of csPCa relies on prostate biopsy, which carries complications including pain, hemorrhage, and infection( 5 – 7 ). A study showed that 30% patients felt specific pain during the biopsy( 6 ). Another study indicated the rate of severe infection could reach 2.1% despite quinolone was used prophylactically( 8 ). Hemorrhage is more common than infection. Although most bleedings are self-limiting, severe complications like hemoperitoneum after biopsy have also been reported( 9 ). Thus, it is meaningful to find a non-invasive way to diagnose csPCa. Many studies have indicated that change of expression levels of genes on the Y chromosome might be associated with PCa ( 10 – 12 ). PRKY, a pseudogene on Y chromosome, was observed to have a lower expression in the PCa tissue( 13 ). Our previous study has reviewed that the down regulation of PRKY might attribute to the high methylation level of CpG sites cg05163709 and cg05618150 on its promotor( 14 ). Notably, elevated cg05163709 methylation level is also observed in the urine sample of PCa patients, which suggested the potential of cg05163709 methylation as a novel liquid biopsy marker for csPCa( 15 ). Magnetic resonance imaging (MRI) is a significant examination for PCa. The Prostate Imaging Reporting and Data System (PI-RADS) scoring system, since its introduction in 2012, has been widely used in clinical practice, and recommended as the standardized reporting system for prostate MRI( 2 , 16 ). However, the score is influenced by the experience and expertise of interpreting radiologists, which might bring considerable subjectivity in its assessment outcomes( 17 ). An computer assisted automated image interpretation systems can effectively address these issues by providing standardized analysis process, and enhance the diagnosis efficiency of PCa. The concept of radiomics is proposed in 2012, which described a subject using image features to construct prediction models for diseases( 18 ). It is now been used in multiple researches in the PCa. In 2015, Khalvati et al. used support vector machine (SVM) algorithm to predict PCa based on MRI features, achieving an AUC of 0.88( 19 ). Another study utilized MR imaging features to predict PCa in patients with prostate specific antigen (PSA) levels of 4–10 ng/mL, in which the radiomics model exhibited superior diagnostic performance compared to the interpretation of radiologists( 20 ). For the diagnosis of csPCa, models constructed by Li and Castillo both reached high predictive value( 21 , 22 ). However, there remains few research integrating radiomics with novel biological markers to predict csPCa, which indicates that while a foundational framework of radiomics studies in csPCa diagnosis has been established, further researches were still needful to enrich the framework. This study aims to evaluate the diagnostic efficacy of PRKY promoter methylation sites, cg05163709 and cg05618150 in urine samples for csPCa detection, and develop an effective prediction model combining MRI radiomics and methylation biomarkers. By this means, we hope to explore a non-invasive diagnostic strategy for csPCa diagnosis to avoid unnecessary prostate biopsy. Method 1.1 Patient Inclusion and Exclusion Criteria This study was approved by the institutional ethics committee of our hospital. Patients hospitalized between January 2022 and January 2023, diagnosed as benign prostatic hyperplasia (BPH) or been suspected as PCa, who underwent prostate biopsy or holmium laser enucleation of the prostate (HoLEP), were included. All participants have signed written informed consents. We collected their clinical data including age, PSA level and prostate volume. The urine samples were obtained after prostate massage. The inclusion and exclusion criteria is shown in Fig. 1 . A total of 39 patients were finally included in our study. 1.2 Acquisition and Analysis of Pathological Data All patients underwent ultrasound-guided systematic prostate biopsy or HoLEP. Tissue specimens were analyzed by the pathology department of our hospital. Gleason score of the malignant specimens were estimated by two pathologists. For patients with multiple malignant cores, the highest Gleason score was recorded. 1.3 Urine Samples Collection and Methylation Analysis Urine samples were collected after prostate massage, and centrifuged for 10 minutes within 4 hours after collection (EppendorfAG, Germany) (centrifugal radius, 9cm; centrifugal force, 4000×g;). DNA was extracted using the QIAGEN QIAamp® DNA Mini Kit, and the the EZ DNA Methylation-Gold™ Kit was used in bisulfite conversion. DNA probes (Zhengze, China) were designed for detecting the methylation level of cg05163709 and cg05618150. ACTB was chosen as control. The recipe of the premix solution is shown in Table 1 . The methylation level of each sample was analyzed via quantitative real-time polymerase chain reaction (qPCR) (initial denaturation, 95℃ for 5 seconds; denaturation, 95℃ for 15 seconds; annealing, extension and fluorescent detection, 56℃ for 1 minutes). The type of the instrument was ABI7500 (Thermo Fisher, America). The difference in cycling threshold (Ct) between the sample and control group (△Ct) was used for quantification of the methylation levels. 1.4 MR images Acquisition MR (Ingenia, Philips) scanning was performed within one month before the surgery or biopsy, including T2WI, DWI (b values = 100,1000 and 2000) maps. The ADC maps were calculated by a designed work station in our hospital. Parameters are listed in Table 2 . All the images were downloaded as DICOM format for analysis. 1.5 Delineation of Regions of Interest ROIs MR images were analyzed by two radiologists with an over-10-year experience. The region of interest (ROI) was delineated on 3D-Slicer, focusing on T2WI hypointense, DWI hyperintense, and ADC hypointense regions (Fig. 2 ). The highest PI-RADS score of each patient was also estimated by the two radiologists. 1.6 Radiomic Feature Extraction Through a plugin, PyRadiomics, of 3D − Slicer, 321 features from the ADC maps (b = 100 − 1000,100 − 2000), and the T2WI maps were extracted. Each series includes 14 shape features, 18 first-order features, and 75 texture features. The texture features include 24 gray level co-occurrence (GLCM) features, 14 neighboring gray level dependence matrix features (NGLDM) features, 16 gray level run length matrix (GLRLM) features,16 gray level size zone matrix (GLSZM) features, 5 neighboring gray tone difference matrix (NGTDM) features. LASSO regression and t-test were employed for feature selection in Python. 1.7 Statistical Analysis Clinical data and methylation levels were analyzed in R 4.2.3. Normality was assessed via Shapiro-Wilk test; parametric t − test or Mann–Whitney test was applied for different type of data. ROC curve was used to evaluated the diagnostic performance. Random forest (RF) models integrating clinical data, methylation levels, and radiomics features were developed in Python, a 3 − fold cross validation was used to test the model. The workflow is summarized in Fig. 3 . Results 2.1 Baseline Characteristics 26 csPCa patients and 13 patients with BPH or cisPCa were enrolled in this study. The mean age of all participants was (72.28 ± 5.54) years. There was no significant difference in age ( P = 0.38), PSA levels ( P = 0.26), or prostate volume ( P = 0.07). However, PI-RADS scores was significantly different ( P = 0.01) between the two groups, which is exhibited in Table 3 . 2.2 PRKY Promoter Methylation Analysis ΔCt values of cg05163709 and cg05618150 were analyzed. No significant difference was observed at cg05618150 ( P = 0.66) between the two groups. In contrast, the csPCa group exhibited significantly lower ΔCt values at cg05163709 compared to the cisPCa & BPH group ( P = 0.0392), indicating higher methylation levels in csPCa patients (Fig. 4 ). The ROC curve demonstrated the predictive efficacy of cg05163709 methylation for csPCa (AUC = 0.75, P = 0.0392). 2.3 MRI Radiomic Feature Selection LASSO regression selected 12 T2WI-derived and 12 ADC-derived radiomic features, including 6 shape features, 3 first-order features, and 15 texture features. The detailed names were shown as follows : T2_shape_Elongation, T2_shape_MajorAxisLength, T2_firstorder_Maximum, T2_firstorder_Minimum, T2_glcm_Imc2, T2_glcm_InverseVariance, T2_gldm_LargeDependenceLowGrayLevelEmphasis, T2_glrlm_RunVariance, T2_glszm_GrayLevelNonUniformityNormalized, T2_glszm_SmallAreaLowGrayLevelEmphasis, T2_ngtdm_Coarseness,T2_ngtdm_Contrast, dADC100_2000_shape_Elongation, dADC100_2000_shape_MajorAxisLength, dADC100_2000_firstorder_Maximum, dADC100_2000_glcm_Idm, dADC100_2000_glcm_InverseVariance, dADC100_2000_ngtdm_Strength, dADC100_1000_shape_Elongation, dADC100_1000_shape_Flatness, dADC100_1000_glcm_Contrast, dADC100_1000_glcm_InverseVariance, dADC100_1000_ngtdm_Contrast, dADC100_1000_glszm_SizeZoneNonUniformityNormalized, which is shown in Fig. 5 . 2.4 Model Development and Validation The performance of the predictive models is exhibited in Fig. 6 . The sensitivity, specificity, accuracy of the T2WI model were 1.0, 0.67 and 0.92; of the ADC model, the T2WI and ADC combined model, the radiomics features and clinical data combined model, as well as the radiomics features, clinical data and PRKY promotor methylation levels combined model were 1.0, 0.80 and 0.92; of the PI-RADS score, clinical data and PRKY promotor methylation levels combined model were 1.0, 0.60 and 0.83. The AUC of the T2WI model was 0.81, of the ADC model as well as the T2WI and ADC combined model was 0.91, of the radiomics features and clinical data combined model was 0.94. The model combined with radiomics features, clinical data and PRKY promotor methylation levels had reached an AUC of 0.97, which was higher than the PI-RADS-based model (AUC = 0.86). Discussion Magnetic resonance imaging (MRI) and ultrasonography are both crucial examinations for PCa. Compared to ultrasonography, MRI provides more objective and comprehensive multiparametric imaging data, which can assist tumor staging assessment( 23 ). However, conventional interpretation of MRI was limited by human visual resolution and subjective variability. Radiomics can addresses these limitations through computerized feature extraction and selection, showing superior objectivity in image analysis. To establish a multidimensional predictive model, radiomics features, clinical data and PRKY promotor methylation levels were all integrated in our study, and the AUC of the combined model had reached 0.97. Our previous research had revealed significantly elevated methylation levels at PRKY promoter CpG sites, cg05618150 and cg05163709, in PCa compared to BPH tissues( 14 ). As Yao’s research had found the association between PCa and the cg05163709 methylation levels in urine samples( 15 ), we subsequently analyzed the connection between methylation levels of cg05163709 as well as cg05618150 and csPCa. The results showed that the cg05163709 methylation level had reached an AUC of 0.74, which was subsequently incorporated into our radiomics model to enhance the predictive value. In radiomics modeling, T2WI and ADC sequences showed superior diagnostic performance (AUC = 0.81 and 0.91). Notably, while DWI maps were considered in delineation, its features were excluded from the final model due to the T2 shine-through effect( 24 ). The ADC sequence, derived from multi-b-value DWI calculations, could effectively solve this limitation( 25 , 26 ), providing more accurate information of tumor-related diffusion restriction. The radiomics-methylation-clinic-combined model showed higher predictive value (AUC = 0.97 vs. 0.86) than the PI-RADS-methylation-clinic combined model, demonstrating the possibly enhanced diagnostic accuracy of computerized texture analysis over conventional visual interpretation. However, due to the small sample size we have, further research is needed to certify this conclusion. To reach the highest interpretability, random forest (RF) was ultimately selected in this study. The RF algorithm, based on the decision trees, provides transparent visualization of the decision processes. It can also accommodate heterogeneous variable types and distributions ( 27 , 28 ). Thus, data including clinical information, MRI features, and methylation levels could be integrated in a combined model. Previous researches combining biomarkers and radiomics mainly focused on conventional parameters such as PSA and PSA density (PSAD) ( 21 , 29 , 30 ). And in Sultan’s study, the OPKO4K score and ExoDx score were also included into the model to enhance the predictive efficiency( 31 ). Distinct from these studies, our radiomics model incorporated PRKY promoter methylation levels as a new biomarker and exhibited a high predictive value. The limitations of our study are as follows. Firstly, this is a single-center research with small sample size and absence of external validation. Secondly, this study is lack of exploration of the mechanism between PRKY methylation-clinical and csPCa. Conclusion In summary, this study investigated the predictive value of PRKY promoter methylation levels in urine samples for csPCa, and established predictive models combining radiomics features with methylation levels in urine samples. This research proposed a novel approach to enhance MRI diagnostic accuracy and diagnose csPCa less invasively. Abbreviations PCa prostate cancer csPCa clinically significant prostate cancer cisPCa clinically insignificant prostate cancer MRI Magnetic resonance imaging ROI Region of interest RF Random forest PI-RADS Prostate Imaging Reporting and Data System HoLEP holmium laser enucleation of the prostate BPH benign prostatic hyperplasia Declarations Ethics approval and consent to participate All experiments in this study was performed in accordance with relevant guidelines and regulations. This study was approved by the Ethics Committees of the Second Affiliated Hospital of Soochow University. The approval number is JDLK202205901 ( date June 20, 2022 ). All the patients were informed about and provided consent for the study, and written informed consent was obtained from each participant . Consent for publication All the listed authors have seen and approved the manuscript for publication. Data availability statement Data is provided within the manuscript or supplementary information files. Financial & competing interests disclosure This work was supported by the National Natural Science Foundation of China (No. 81773221), the Suzhou Gusu Health Talents Research Project (GSWS2021016), the SuZhou Municipal Science and Technology Project (SS201857). The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. No writing assistance was utilized in the production of this manuscript. Author contributions Z Dai, H Chen, Y Zhou, and J Zhu conceived and designed the study; Z Dai, Y Wang, Z Chen, G Xu, S Sun, W Liu, G Yang, Z Zhou, B Li, X Wang and X Wang analyzed the data and prepared the figures; Z Dai, Y Wang, and Z Chen wrote and revised the manuscript. All authors have read and agreed to the final version of the manuscript. Acknowledgements The authors are grateful for the support from the Second Affiliated Hospital of Soochow University and the Third Affiliated Hospital of Anhui Medical University. References Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–49. Cornford P, van den Bergh RCN, Briers E, Van den Broeck T, Brunckhorst O, Darraugh J, et al. EAU-EANM-ESTRO-ESUR-ISUP-SIOG Guidelines on Prostate Cancer-2024 Update. Part I: Screening, Diagnosis, and Local Treatment with Curative Intent. Eur Urol. 2024;86(2):148–63. Labbate CV, Paner GP, Eggener SE. Should Grade Group 1 (GG1) be called cancer? World J Urol. 2022;40(1):15–9. Palsdottir T, Nordström T, Aly M, Lindberg J, Clements M, Egevad L, et al. Are Prostate Specific-Antigen (PSA) and age associated with the risk of ISUP Grade 1 prostate cancer? Results from 72 996 individual biopsy cores in 6 083 men from the Stockholm3 study. PloS One. 2019;14(6):e0218280. Borghesi M, Ahmed H, Nam R, Schaeffer E, Schiavina R, Taneja S, et al. Complications After Systematic, Random, and Image-guided Prostate Biopsy. Eur Urol. 2017 Mar;71(3):353–65. Collins GN, Lloyd SN, Hehir M, McKelvie GB. Multiple transrectal ultrasound-guided prostatic biopsies--true morbidity and patient acceptance. Br J Urol. 1993;71(4):460–3. Irani J, Fournier F, Bon D, Gremmo E, Doré B, Aubert J. Patient tolerance of transrectal ultrasound-guided biopsy of the prostate. Br J Urol. 1997;79(4):608–10. Ozden E, Bostanci Y, Yakupoglu KY, Akdeniz E, Yilmaz AF, Tulek N, et al. Incidence of acute prostatitis caused by extended-spectrum beta-lactamase-producing Escherichia coli after transrectal prostate biopsy. Urology. 2009;74(1):119–23. Frascheri MF, Contreras P, Blas L, Bonanno N, Ameri C. Hemoperitoneum after transperineal prostate biopsy. Medicina (Mex). 2022;82(3):452–5. Lindström S, Adami HO, Adolfsson J, Wiklund F. Y chromosome haplotypes and prostate cancer in Sweden. Clin Cancer Res Off J Am Assoc Cancer Res. 2008;14(20):6712–6. Lau YF, Zhang J. Expression analysis of thirty one Y chromosome genes in human prostate cancer. Mol Carcinog. 2000;27(4):308–21. Yadav SK, Kumari A, Javed S, Ali S. DYZ1 arrays show sequence variation between the monozygotic males. BMC Genet. 2014;15:19. Dasari VK, Goharderakhshan RZ, Perinchery G, Li LC, Tanaka Y, Alonzo J, et al. Expression analysis of Y chromosome genes in human prostate cancer. J Urol. 2001;165(4):1335–41. Dai Z, Chen H, Feng K, Li T, Liu W, Zhou Y, et al. Promoter hypermethylation of Y-chromosome gene PRKY as a potential biomarker for the early diagnosis of prostate cancer. Epigenomics. 2024;16(11–12):835–50. Yao L, Ren S, Zhang M, Du F, Zhu Y, Yu H, et al. Identification of specific DNA methylation sites on the Y-chromosome as biomarker in prostate cancer. Oncotarget. 2015;6(38):40611–21. Barentsz JO, Richenberg J, Clements R, Choyke P, Verma S, Villeirs G, et al. ESUR prostate MR guidelines 2012. Eur Radiol. 2012;22(4):746–57. Forookhi A, Laschena L, Pecoraro M, Borrelli A, Massaro M, Dehghanpour A, et al. Bridging the experience gap in prostate multiparametric magnetic resonance imaging using artificial intelligence: A prospective multi-reader comparison study on inter-reader agreement in PI-RADS v2.1, image quality and reporting time between novice and expert readers. Eur J Radiol. 2023;161:110749. Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RGPM, Granton P, et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer Oxf Engl 1990. 2012;48(4):441–6. Khalvati F, Wong A, Haider MA. Automated prostate cancer detection via comprehensive multi-parametric magnetic resonance imaging texture feature models. BMC Med Imaging. 2015;15:27. Zhong JG, Shi L, Liu J, Cao F, Ma YQ, Zhang Y. Predicting prostate cancer in men with PSA levels of 4-10 ng/mL: MRI-based radiomics can help junior radiologists improve the diagnostic performance. Sci Rep. 2023;13(1):4846. Li M, Chen T, Zhao W, Wei C, Li X, Duan S, et al. Radiomics prediction model for the improved diagnosis of clinically significant prostate cancer on biparametric MRI. Quant Imaging Med Surg. 2020;10(2):368–79. Castillo T JM, Arif M, Starmans MPA, Niessen WJ, Bangma CH, Schoots IG, et al. Classification of Clinically Significant Prostate Cancer on Multi-Parametric MRI: A Validation Study Comparing Deep Learning and Radiomics. Cancers. 2021;14(1):12. Penzkofer T, Tempany-Afdhal CM. Prostate Cancer Detection and Diagnosis: The Role of MR and its Comparison to other Diagnostic Modalities – A Radiologist’s Perspective. NMR Biomed. 2014;27(1):10.1002/nbm.3002. Duran R, Ronot M, Kerbaol A, Van Beers B, Vilgrain V. Hepatic hemangiomas: factors associated with T2 shine-through effect on diffusion-weighted MR sequences. Eur J Radiol. 2014;83(3):468–78. Teică RV, Șerbănescu MS, Florescu LM, Gheonea IA. Tumor Area Highlighting Using T2WI, ADC Map, and DWI Sequence Fusion on bpMRI Images for Better Prostate Cancer Diagnosis. Life. 2023;13(4):910. Park SY, Kim CK, Park JJ, Park BK. Exponential apparent diffusion coefficient in evaluating prostate cancer at 3 T: preliminary experience. Br J Radiol. 2016;89(1058):20150470. Erickson BJ, Korfiatis P, Akkus Z, Kline TL. Machine Learning for Medical Imaging. Radiogr Rev Publ Radiol Soc N Am Inc. 2017;37(2):505–15. Breiman L. Random Forests. Mach Learn. 2001;45(1):5–32. Krauss W, Frey J, Heydorn Lagerlöf J, Lidén M, Thunberg P. Radiomics from multisite MRI and clinical data to predict clinically significant prostate cancer. Acta Radiol Stockh Swed 1987. 2024;65(3):307–17. Gaudiano C, Mottola M, Bianchi L, Corcioni B, Cattabriga A, Cocozza MA, et al. Beyond Multiparametric MRI and towards Radiomics to Detect Prostate Cancer: A Machine Learning Model to Predict Clinically Significant Lesions. Cancers. 2022;14(24):6156. Sultan MI, Huynh LM, Kamil S, Abdelaziz A, Hammad MA, Gin GE, et al. Utility of noninvasive biomarker testing and MRI to predict a prostate cancer diagnosis. Int Urol Nephrol. 2023;56(2):539. Tables Tables 1-3 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table1.xlsx Table3.xlsx Table2.xlsx originalmaterials.xlsx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 06 Aug, 2025 Reviewers agreed at journal 30 Jul, 2025 Reviewers invited by journal 18 Jul, 2025 Editor assigned by journal 13 Jun, 2025 Editor invited by journal 09 Jun, 2025 Submission checks completed at journal 08 Jun, 2025 First submitted to journal 08 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6806278","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":487477312,"identity":"696a1be4-bc34-4b68-8e12-2a54581ab24f","order_by":0,"name":"Zheng Dai","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYBAC+/bmgw8+/rGpb2NvIFKLAc+xZMOZDWmM/TwHiNUi4WMmzdlwmHHmjAQitZhLsCUbM+5IYza4+XjjDYYam2iCWixnNx98XHjGhs3gdlqxBcOxtNwGgnruHEs2nsGWxmNwO8dMgrHhMBFabuSYSfOwHZYwuHmGSC0GIC28bYcNJGfwEKlFsgcYyDPOpCXw8wD9kkCMX/jZgVH5ocImgY398MYbH2psiPALsiMlEkhRDtFCqo5RMApGwSgYGQAAf71EnZFHAe0AAAAASUVORK5CYII=","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":true,"prefix":"","firstName":"Zheng","middleName":"","lastName":"Dai","suffix":""},{"id":487477313,"identity":"f7145c98-a51a-4c8f-9a1e-51f453166074","order_by":1,"name":"Yufei Wang","email":"","orcid":"","institution":"The Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Yufei","middleName":"","lastName":"Wang","suffix":""},{"id":487477314,"identity":"84fa2b7a-fdcf-4ad3-9056-fbd8a8f47a52","order_by":2,"name":"Zeyu Chen","email":"","orcid":"","institution":"The Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Zeyu","middleName":"","lastName":"Chen","suffix":""},{"id":487477315,"identity":"3059f923-36bf-403f-9b68-6f03bbadd46d","order_by":3,"name":"Guosheng Xu","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Guosheng","middleName":"","lastName":"Xu","suffix":""},{"id":487477316,"identity":"5729f6a8-d964-4551-848c-c9e9f8714196","order_by":4,"name":"Song Sun","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Song","middleName":"","lastName":"Sun","suffix":""},{"id":487477317,"identity":"2855ee4a-828d-4123-872a-403878b3585b","order_by":5,"name":"Wei Liu","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Liu","suffix":""},{"id":487477318,"identity":"1490b821-0d70-416f-ad03-02f270de40dc","order_by":6,"name":"Guang Yang","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Guang","middleName":"","lastName":"Yang","suffix":""},{"id":487477319,"identity":"dd488763-6b0b-4feb-8df3-1c1fa7f72491","order_by":7,"name":"Zhengxing Zhou","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zhengxing","middleName":"","lastName":"Zhou","suffix":""},{"id":487477320,"identity":"fe0c9777-f501-42f3-a9e0-59bfbfea2343","order_by":8,"name":"Bo Li","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Li","suffix":""},{"id":487477321,"identity":"6d62effe-3870-4dbb-8c69-44ecc41d0b54","order_by":9,"name":"Xuchang Wang","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xuchang","middleName":"","lastName":"Wang","suffix":""},{"id":487477324,"identity":"acee5ee9-394a-408b-9d77-f3fe7fe91f19","order_by":10,"name":"Yibin Zhou","email":"","orcid":"","institution":"The Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Yibin","middleName":"","lastName":"Zhou","suffix":""},{"id":487477325,"identity":"821dd6a7-f8f2-4a1b-82b4-abd2607c3f4f","order_by":11,"name":"Jin Zhu","email":"","orcid":"","institution":"The Second Affiliated Hospital of Soochow University","correspondingAuthor":false,"prefix":"","firstName":"Jin","middleName":"","lastName":"Zhu","suffix":""},{"id":487477326,"identity":"07e245fa-ef13-476c-91f4-d3e21d93cbfb","order_by":12,"name":"Hongbing Chen","email":"","orcid":"","institution":"The Third Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hongbing","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2025-06-03 02:53:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6806278/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6806278/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87386790,"identity":"b5c448ac-0ab6-4da6-a534-e888d6e3372f","added_by":"auto","created_at":"2025-07-23 09:00:31","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":414110,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInclusion and exclusion criteria.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Figure1.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/49ff9020e1dbd22dec18dcdf.jpg"},{"id":87385186,"identity":"8f0e1084-ac1c-4431-8f7c-f0c912e93bf3","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1449810,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eROI delineation.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRegions which showed hypointense in T2WI maps, hyperintense in DWI maps and hypointense in ADC maps was delineated. The color of the DWI images were reversed in our hospital.\u003c/p\u003e","description":"","filename":"Figure2.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/20c9a2af5a8bbfd3039a1014.jpg"},{"id":87386791,"identity":"72f5fbbc-746b-4ded-99c3-49e27815518f","added_by":"auto","created_at":"2025-07-23 09:00:31","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":461295,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe work flow of our study.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Figure3.tif.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/2e876ae5097a25f632245fe2.jpg"},{"id":87387788,"identity":"d679cfbf-68b8-4e97-8a53-6a64ed6bf962","added_by":"auto","created_at":"2025-07-23 09:08:31","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":790108,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of cg05163709 methylation levels between csPCa and ncsPCa or BPH patients, as well as the ROC curve for cg05163709 methylation level to predict csPCa.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) TheΔCt of cg05163709 in csPCa patients was higher than which in ncsPCa or BPH patients, which indicated that the csPCa group showed a higher methylation level. (B) The AUC of cg05163709 methylation level was 0.74。\u003c/p\u003e","description":"","filename":"Figure4.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/8ffd7badd81a11c1459217de.jpg"},{"id":87385195,"identity":"5ae9fc81-e482-444a-af86-361cc6634654","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":360358,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSelected radiomics features.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe\u003cstrong\u003e \u003c/strong\u003edots with warm colors refer to positive correlation with csPCa, while the dots with cold colors refer to negative correlation.\u003c/p\u003e","description":"","filename":"Figure5.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/41a406752462245314af2966.jpg"},{"id":87385190,"identity":"c2d5d817-ff87-45ba-9ed5-2ac9737ea6bf","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":497586,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eROC curves of the predictive models.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) The ROC curve of T2WI-based model; (B) The ROC curve of ADC-based model; (C) The ROC curve of T2WI-ADC-combined model; (D) The ROC curve of T2WI-ADC-clinic-combined model; (E) The ROC curve of T2WI-ADC-clinic-methylation-combined model; (F)\u003c/p\u003e\n\u003cp\u003eThe ROC curve of PI-RADS-clinic-methylation-combined model.\u003c/p\u003e","description":"","filename":"Figure6.tiff.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/7b7568427f5f7b8b0983f41c.jpg"},{"id":87467131,"identity":"2833ce6c-53b3-466e-b885-8f0d2678b35e","added_by":"auto","created_at":"2025-07-24 08:00:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4804453,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/99026e35-4a2f-4029-8c18-e9b15255ae67.pdf"},{"id":87385182,"identity":"99d7969d-a1b6-4dd1-8742-f4c6d95401e3","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":10180,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/ca90bfa9ff56d9a62485b03e.xlsx"},{"id":87386792,"identity":"9bbfd885-67d1-4c18-86cb-22761e957ac4","added_by":"auto","created_at":"2025-07-23 09:00:31","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":11018,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/dfdccba0d7543c98fb8b14ad.xlsx"},{"id":87385189,"identity":"a9aa2d8f-8b5b-43e0-bc8a-149098c7f8ce","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":10721,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/cc9ae28560008b459c137021.xlsx"},{"id":87385194,"identity":"4f73bed9-c822-448f-a6fd-9bc21e398c64","added_by":"auto","created_at":"2025-07-23 08:52:31","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":170380,"visible":true,"origin":"","legend":"","description":"","filename":"originalmaterials.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6806278/v1/9f057c5c0d7c5f88fbb61c36.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Constructing a Prediction Model for Clinically Significant Prostate Cancer Combined with Radiomics Features of MRI and PRKY Promoter Methylation Level in Urine Samples","fulltext":[{"header":"Introduction","content":"\u003cp\u003eProstate cancer (PCa) is one of the most common cancer in males and the leading cause of cancer-related mortality among men in many countries(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Due to its insidious symptoms, many patients are diagnosed at advanced stages. Thus, early and accurate diagnosis of PCa is crucial. Pathological examination of prostate tissue is the gold standard for PCa diagnosis(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). According to the International Society of Urological Pathology (ISUP) grading system, PCa is classified into grades 1 to 5 based on tumor features. The 2024 European Association of Urology (EAU) guidelines defined ISUP grade 1 (Gleason score 3\u0026thinsp;+\u0026thinsp;3\u0026thinsp;=\u0026thinsp;6) as clinically insignificant prostate cancer (cisPCa), while tumors with ISUP grades\u0026thinsp;\u0026ge;\u0026thinsp;2 were classified as clinically significant prostate cancer (csPCa)(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). These two categories exhibit differences in prognosis. The EAU guidelines suggest that cisPCa showes better prognosis than csPCa, active surveillance should be considered firstly rather than early intervention. In contrast, csPCa patients require prompt surgery or endocrine therapy once diagnosed(\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDiagnosis of csPCa relies on prostate biopsy, which carries complications including pain, hemorrhage, and infection(\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). A study showed that 30% patients felt specific pain during the biopsy(\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Another study indicated the rate of severe infection could reach 2.1% despite quinolone was used prophylactically(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Hemorrhage is more common than infection. Although most bleedings are self-limiting, severe complications like hemoperitoneum after biopsy have also been reported(\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). Thus, it is meaningful to find a non-invasive way to diagnose csPCa.\u003c/p\u003e\u003cp\u003eMany studies have indicated that change of expression levels of genes on the Y chromosome might be associated with PCa (\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). PRKY, a pseudogene on Y chromosome, was observed to have a lower expression in the PCa tissue(\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Our previous study has reviewed that the down regulation of PRKY might attribute to the high methylation level of CpG sites cg05163709 and cg05618150 on its promotor(\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Notably, elevated cg05163709 methylation level is also observed in the urine sample of PCa patients, which suggested the potential of cg05163709 methylation as a novel liquid biopsy marker for csPCa(\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eMagnetic resonance imaging (MRI) is a significant examination for PCa. The Prostate Imaging Reporting and Data System (PI-RADS) scoring system, since its introduction in 2012, has been widely used in clinical practice, and recommended as the standardized reporting system for prostate MRI(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). However, the score is influenced by the experience and expertise of interpreting radiologists, which might bring considerable subjectivity in its assessment outcomes(\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). An computer assisted automated image interpretation systems can effectively address these issues by providing standardized analysis process, and enhance the diagnosis efficiency of PCa.\u003c/p\u003e\u003cp\u003eThe concept of radiomics is proposed in 2012, which described a subject using image features to construct prediction models for diseases(\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). It is now been used in multiple researches in the PCa. In 2015, Khalvati et al. used support vector machine (SVM) algorithm to predict PCa based on MRI features, achieving an AUC of 0.88(\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). Another study utilized MR imaging features to predict PCa in patients with prostate specific antigen (PSA) levels of 4\u0026ndash;10 ng/mL, in which the radiomics model exhibited superior diagnostic performance compared to the interpretation of radiologists(\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). For the diagnosis of csPCa, models constructed by Li and Castillo both reached high predictive value(\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e). However, there remains few research integrating radiomics with novel biological markers to predict csPCa, which indicates that while a foundational framework of radiomics studies in csPCa diagnosis has been established, further researches were still needful to enrich the framework.\u003c/p\u003e\u003cp\u003eThis study aims to evaluate the diagnostic efficacy of PRKY promoter methylation sites, cg05163709 and cg05618150 in urine samples for csPCa detection, and develop an effective prediction model combining MRI radiomics and methylation biomarkers. By this means, we hope to explore a non-invasive diagnostic strategy for csPCa diagnosis to avoid unnecessary prostate biopsy.\u003c/p\u003e"},{"header":"Method","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e1.1 Patient Inclusion and Exclusion Criteria\u003c/h2\u003e\u003cp\u003eThis study was approved by the institutional ethics committee of our hospital. Patients hospitalized between January 2022 and January 2023, diagnosed as benign prostatic hyperplasia (BPH) or been suspected as PCa, who underwent prostate biopsy or holmium laser enucleation of the prostate (HoLEP), were included. All participants have signed written informed consents. We collected their clinical data including age, PSA level and prostate volume. The urine samples were obtained after prostate massage. The inclusion and exclusion criteria is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. A total of 39 patients were finally included in our study.\u003c/p\u003e\n\u003ch3\u003e1.2 Acquisition and Analysis of Pathological Data\u003c/h3\u003e\n\u003cp\u003eAll patients underwent ultrasound-guided systematic prostate biopsy or HoLEP. Tissue specimens were analyzed by the pathology department of our hospital. Gleason score of the malignant specimens were estimated by two pathologists. For patients with multiple malignant cores, the highest Gleason score was recorded.\u003c/p\u003e\n\u003ch3\u003e1.3 Urine Samples Collection and Methylation Analysis\u003c/h3\u003e\n\u003cp\u003eUrine samples were collected after prostate massage, and centrifuged for 10 minutes within 4 hours after collection (EppendorfAG, Germany) (centrifugal radius, 9cm; centrifugal force, 4000\u0026times;g;). DNA was extracted using the QIAGEN QIAamp\u0026reg; DNA Mini Kit, and the the EZ DNA Methylation-Gold\u0026trade; Kit was used in bisulfite conversion. DNA probes (Zhengze, China) were designed for detecting the methylation level of cg05163709 and cg05618150. ACTB was chosen as control. The recipe of the premix solution is shown in \u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e. The methylation level of each sample was analyzed via quantitative real-time polymerase chain reaction (qPCR) (initial denaturation, 95℃ for 5 seconds; denaturation, 95℃ for 15 seconds; annealing, extension and fluorescent detection, 56℃ for 1 minutes). The type of the instrument was ABI7500 (Thermo Fisher, America). The difference in cycling threshold (Ct) between the sample and control group (△Ct) was used for quantification of the methylation levels.\u003c/p\u003e\n\u003ch3\u003e1.4 MR images Acquisition\u003c/h3\u003e\n\u003cp\u003eMR (Ingenia, Philips) scanning was performed within one month before the surgery or biopsy, including T2WI, DWI (b values\u0026thinsp;=\u0026thinsp;100,1000 and 2000) maps. The ADC maps were calculated by a designed work station in our hospital. Parameters are listed in \u003cb\u003eTable\u0026nbsp;2\u003c/b\u003e. All the images were downloaded as DICOM format for analysis.\u003c/p\u003e\n\u003ch3\u003e1.5 Delineation of Regions of Interest ROIs\u003c/h3\u003e\n\u003cp\u003eMR images were analyzed by two radiologists with an over-10-year experience. The region of interest (ROI) was delineated on 3D-Slicer, focusing on T2WI hypointense, DWI hyperintense, and ADC hypointense regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The highest PI-RADS score of each patient was also estimated by the two radiologists.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e1.6 Radiomic Feature Extraction\u003c/h2\u003e\u003cp\u003eThrough a plugin, PyRadiomics, of 3D\u0026thinsp;\u0026minus;\u0026thinsp;Slicer, 321 features from the ADC maps (b\u0026thinsp;=\u0026thinsp;100\u0026thinsp;\u0026minus;\u0026thinsp;1000,100\u0026thinsp;\u0026minus;\u0026thinsp;2000), and the T2WI maps were extracted. Each series includes 14 shape features, 18 first-order features, and 75 texture features. The texture features include 24 gray level co-occurrence (GLCM) features, 14 neighboring gray level dependence matrix features (NGLDM) features, 16 gray level run length matrix (GLRLM) features,16 gray level size zone matrix (GLSZM) features, 5 neighboring gray tone difference matrix (NGTDM) features. LASSO regression and t-test were employed for feature selection in Python.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003e1.7 Statistical Analysis\u003c/h3\u003e\n\u003cp\u003eClinical data and methylation levels were analyzed in R 4.2.3. Normality was assessed via Shapiro-Wilk test; parametric t\u0026thinsp;\u0026minus;\u0026thinsp;test or Mann\u0026ndash;Whitney test was applied for different type of data. ROC curve was used to evaluated the diagnostic performance. Random forest (RF) models integrating clinical data, methylation levels, and radiomics features were developed in Python, a 3\u0026thinsp;\u0026minus;\u0026thinsp;fold cross validation was used to test the model. The workflow is summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Baseline Characteristics\u003c/h2\u003e\u003cp\u003e26 csPCa patients and 13 patients with BPH or cisPCa were enrolled in this study. The mean age of all participants was (72.28\u0026thinsp;\u0026plusmn;\u0026thinsp;5.54) years. There was no significant difference in age (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.38), PSA levels (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.26), or prostate volume (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.07). However, PI-RADS scores was significantly different (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.01) between the two groups, which is exhibited in \u003cb\u003eTable\u0026nbsp;3\u003c/b\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e2.2 PRKY Promoter Methylation Analysis\u003c/h2\u003e\u003cp\u003eΔCt values of cg05163709 and cg05618150 were analyzed. No significant difference was observed at cg05618150 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.66) between the two groups. In contrast, the csPCa group exhibited significantly lower ΔCt values at cg05163709 compared to the cisPCa \u0026amp; BPH group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0392), indicating higher methylation levels in csPCa patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The ROC curve demonstrated the predictive efficacy of cg05163709 methylation for csPCa (AUC\u0026thinsp;=\u0026thinsp;0.75, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0392).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e2.3 MRI Radiomic Feature Selection\u003c/h2\u003e\u003cp\u003eLASSO regression selected 12 T2WI-derived and 12 ADC-derived radiomic features, including 6 shape features, 3 first-order features, and 15 texture features. The detailed names were shown as follows : T2_shape_Elongation, T2_shape_MajorAxisLength, T2_firstorder_Maximum, T2_firstorder_Minimum, T2_glcm_Imc2, T2_glcm_InverseVariance, T2_gldm_LargeDependenceLowGrayLevelEmphasis, T2_glrlm_RunVariance, T2_glszm_GrayLevelNonUniformityNormalized, T2_glszm_SmallAreaLowGrayLevelEmphasis, T2_ngtdm_Coarseness,T2_ngtdm_Contrast, dADC100_2000_shape_Elongation, dADC100_2000_shape_MajorAxisLength, dADC100_2000_firstorder_Maximum, dADC100_2000_glcm_Idm, dADC100_2000_glcm_InverseVariance, dADC100_2000_ngtdm_Strength, dADC100_1000_shape_Elongation, dADC100_1000_shape_Flatness, dADC100_1000_glcm_Contrast, dADC100_1000_glcm_InverseVariance, dADC100_1000_ngtdm_Contrast, dADC100_1000_glszm_SizeZoneNonUniformityNormalized, which is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Model Development and Validation\u003c/h2\u003e\u003cp\u003eThe performance of the predictive models is exhibited in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. The sensitivity, specificity, accuracy of the T2WI model were 1.0, 0.67 and 0.92; of the ADC model, the T2WI and ADC combined model, the radiomics features and clinical data combined model, as well as the radiomics features, clinical data and PRKY promotor methylation levels combined model were 1.0, 0.80 and 0.92; of the PI-RADS score, clinical data and PRKY promotor methylation levels combined model were 1.0, 0.60 and 0.83. The AUC of the T2WI model was 0.81, of the ADC model as well as the T2WI and ADC combined model was 0.91, of the radiomics features and clinical data combined model was 0.94. The model combined with radiomics features, clinical data and PRKY promotor methylation levels had reached an AUC of 0.97, which was higher than the PI-RADS-based model (AUC\u0026thinsp;=\u0026thinsp;0.86).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eMagnetic resonance imaging (MRI) and ultrasonography are both crucial examinations for PCa. Compared to ultrasonography, MRI provides more objective and comprehensive multiparametric imaging data, which can assist tumor staging assessment(\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). However, conventional interpretation of MRI was limited by human visual resolution and subjective variability. Radiomics can addresses these limitations through computerized feature extraction and selection, showing superior objectivity in image analysis. To establish a multidimensional predictive model, radiomics features, clinical data and PRKY promotor methylation levels were all integrated in our study, and the AUC of the combined model had reached 0.97.\u003c/p\u003e\u003cp\u003eOur previous research had revealed significantly elevated methylation levels at PRKY promoter CpG sites, cg05618150 and cg05163709, in PCa compared to BPH tissues(\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). As Yao\u0026rsquo;s research had found the association between PCa and the cg05163709 methylation levels in urine samples(\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e), we subsequently analyzed the connection between methylation levels of cg05163709 as well as cg05618150 and csPCa. The results showed that the cg05163709 methylation level had reached an AUC of 0.74, which was subsequently incorporated into our radiomics model to enhance the predictive value.\u003c/p\u003e\u003cp\u003eIn radiomics modeling, T2WI and ADC sequences showed superior diagnostic performance (AUC\u0026thinsp;=\u0026thinsp;0.81 and 0.91). Notably, while DWI maps were considered in delineation, its features were excluded from the final model due to the T2 shine-through effect(\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). The ADC sequence, derived from multi-b-value DWI calculations, could effectively solve this limitation(\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e), providing more accurate information of tumor-related diffusion restriction.\u003c/p\u003e\u003cp\u003eThe radiomics-methylation-clinic-combined model showed higher predictive value (AUC\u0026thinsp;=\u0026thinsp;0.97 vs. 0.86) than the PI-RADS-methylation-clinic combined model, demonstrating the possibly enhanced diagnostic accuracy of computerized texture analysis over conventional visual interpretation. However, due to the small sample size we have, further research is needed to certify this conclusion.\u003c/p\u003e\u003cp\u003eTo reach the highest interpretability, random forest (RF) was ultimately selected in this study. The RF algorithm, based on the decision trees, provides transparent visualization of the decision processes. It can also accommodate heterogeneous variable types and distributions (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e). Thus, data including clinical information, MRI features, and methylation levels could be integrated in a combined model.\u003c/p\u003e\u003cp\u003ePrevious researches combining biomarkers and radiomics mainly focused on conventional parameters such as PSA and PSA density (PSAD) (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). And in Sultan\u0026rsquo;s study, the OPKO4K score and ExoDx score were also included into the model to enhance the predictive efficiency(\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Distinct from these studies, our radiomics model incorporated PRKY promoter methylation levels as a new biomarker and exhibited a high predictive value.\u003c/p\u003e\u003cp\u003eThe limitations of our study are as follows. Firstly, this is a single-center research with small sample size and absence of external validation. Secondly, this study is lack of exploration of the mechanism between PRKY methylation-clinical and csPCa.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, this study investigated the predictive value of PRKY promoter methylation levels in urine samples for csPCa, and established predictive models combining radiomics features with methylation levels in urine samples. This research proposed a novel approach to enhance MRI diagnostic accuracy and diagnose csPCa less invasively.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePCa\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eprostate cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ecsPCa\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eclinically significant prostate cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ecisPCa\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eclinically insignificant prostate cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMRI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMagnetic resonance imaging\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eROI\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eRegion of interest\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eRF\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eRandom forest\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePI-RADS\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eProstate Imaging Reporting and Data System\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eHoLEP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eholmium laser enucleation of the prostate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eBPH\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ebenign prostatic hyperplasia\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll experiments in this study was performed in accordance with relevant guidelines and regulations. This study was approved by the Ethics Committees of the Second Affiliated Hospital of Soochow University. The approval number is JDLK202205901 (\u003cem\u003edate\u003c/em\u003e June 20, \u003cem\u003e2022\u003c/em\u003e). All the patients were informed about and provided consent for the study, and\u0026nbsp;\u003cem\u003ewritten\u003c/em\u003e \u003cem\u003einformed\u003c/em\u003e \u003cem\u003econsent\u003c/em\u003e was \u003cem\u003eobtained\u003c/em\u003e \u003cem\u003efrom\u003c/em\u003e \u003cem\u003eeach\u003c/em\u003e \u003cem\u003eparticipant\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll the listed authors have seen and approved the manuscript for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData is provided within the manuscript or supplementary information files.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFinancial \u0026amp; competing interests disclosure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (No. 81773221), the\u0026nbsp;Suzhou Gusu Health Talents Research Project\u0026nbsp;(GSWS2021016), the\u0026nbsp;SuZhou Municipal Science and Technology Project\u0026nbsp;(SS201857). The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. No writing assistance was utilized in the production of this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eZ Dai, H Chen, Y Zhou, and J Zhu conceived and designed the study; Z Dai, Y Wang, Z Chen, G Xu, S Sun, W Liu, G Yang, Z Zhou, B Li, X Wang and X Wang analyzed the data and prepared the figures; Z Dai, Y Wang, and Z Chen wrote and revised the manuscript.\u0026nbsp;All authors have read and agreed to the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors are grateful for the support from the Second Affiliated Hospital of Soochow University and the Third Affiliated Hospital of Anhui Medical University.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209\u0026ndash;49. \u003c/li\u003e\n\u003cli\u003eCornford P, van den Bergh RCN, Briers E, Van den Broeck T, Brunckhorst O, Darraugh J, et al. EAU-EANM-ESTRO-ESUR-ISUP-SIOG Guidelines on Prostate Cancer-2024 Update. Part I: Screening, Diagnosis, and Local Treatment with Curative Intent. Eur Urol. 2024;86(2):148\u0026ndash;63. \u003c/li\u003e\n\u003cli\u003eLabbate CV, Paner GP, Eggener SE. Should Grade Group 1 (GG1) be called cancer? World J Urol. 2022;40(1):15\u0026ndash;9. \u003c/li\u003e\n\u003cli\u003ePalsdottir T, Nordstr\u0026ouml;m T, Aly M, Lindberg J, Clements M, Egevad L, et al. Are Prostate Specific-Antigen (PSA) and age associated with the risk of ISUP Grade 1 prostate cancer? Results from 72 996 individual biopsy cores in 6 083 men from the Stockholm3 study. PloS One. 2019;14(6):e0218280. \u003c/li\u003e\n\u003cli\u003eBorghesi M, Ahmed H, Nam R, Schaeffer E, Schiavina R, Taneja S, et al. Complications After Systematic, Random, and Image-guided Prostate Biopsy. Eur Urol. 2017 Mar;71(3):353\u0026ndash;65. \u003c/li\u003e\n\u003cli\u003eCollins GN, Lloyd SN, Hehir M, McKelvie GB. Multiple transrectal ultrasound-guided prostatic biopsies--true morbidity and patient acceptance. Br J Urol. 1993;71(4):460\u0026ndash;3. \u003c/li\u003e\n\u003cli\u003eIrani J, Fournier F, Bon D, Gremmo E, Dor\u0026eacute; B, Aubert J. Patient tolerance of transrectal ultrasound-guided biopsy of the prostate. Br J Urol. 1997;79(4):608\u0026ndash;10. \u003c/li\u003e\n\u003cli\u003eOzden E, Bostanci Y, Yakupoglu KY, Akdeniz E, Yilmaz AF, Tulek N, et al. Incidence of acute prostatitis caused by extended-spectrum beta-lactamase-producing Escherichia coli after transrectal prostate biopsy. Urology. 2009;74(1):119\u0026ndash;23. \u003c/li\u003e\n\u003cli\u003eFrascheri MF, Contreras P, Blas L, Bonanno N, Ameri C. Hemoperitoneum after transperineal prostate biopsy. Medicina (Mex). 2022;82(3):452\u0026ndash;5. \u003c/li\u003e\n\u003cli\u003eLindstr\u0026ouml;m S, Adami HO, Adolfsson J, Wiklund F. Y chromosome haplotypes and prostate cancer in Sweden. Clin Cancer Res Off J Am Assoc Cancer Res. 2008;14(20):6712\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eLau YF, Zhang J. Expression analysis of thirty one Y chromosome genes in human prostate cancer. Mol Carcinog. 2000;27(4):308\u0026ndash;21. \u003c/li\u003e\n\u003cli\u003eYadav SK, Kumari A, Javed S, Ali S. DYZ1 arrays show sequence variation between the monozygotic males. BMC Genet. 2014;15:19. \u003c/li\u003e\n\u003cli\u003eDasari VK, Goharderakhshan RZ, Perinchery G, Li LC, Tanaka Y, Alonzo J, et al. Expression analysis of Y chromosome genes in human prostate cancer. J Urol. 2001;165(4):1335\u0026ndash;41. \u003c/li\u003e\n\u003cli\u003eDai Z, Chen H, Feng K, Li T, Liu W, Zhou Y, et al. Promoter hypermethylation of Y-chromosome gene PRKY as a potential biomarker for the early diagnosis of prostate cancer. Epigenomics. 2024;16(11\u0026ndash;12):835\u0026ndash;50. \u003c/li\u003e\n\u003cli\u003eYao L, Ren S, Zhang M, Du F, Zhu Y, Yu H, et al. Identification of specific DNA methylation sites on the Y-chromosome as biomarker in prostate cancer. Oncotarget. 2015;6(38):40611\u0026ndash;21. \u003c/li\u003e\n\u003cli\u003eBarentsz JO, Richenberg J, Clements R, Choyke P, Verma S, Villeirs G, et al. ESUR prostate MR guidelines 2012. Eur Radiol. 2012;22(4):746\u0026ndash;57. \u003c/li\u003e\n\u003cli\u003eForookhi A, Laschena L, Pecoraro M, Borrelli A, Massaro M, Dehghanpour A, et al. Bridging the experience gap in prostate multiparametric magnetic resonance imaging using artificial intelligence: A prospective multi-reader comparison study on inter-reader agreement in PI-RADS v2.1, image quality and reporting time between novice and expert readers. Eur J Radiol. 2023;161:110749. \u003c/li\u003e\n\u003cli\u003eLambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RGPM, Granton P, et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer Oxf Engl 1990. 2012;48(4):441\u0026ndash;6. \u003c/li\u003e\n\u003cli\u003eKhalvati F, Wong A, Haider MA. Automated prostate cancer detection via comprehensive multi-parametric magnetic resonance imaging texture feature models. BMC Med Imaging. 2015;15:27. \u003c/li\u003e\n\u003cli\u003eZhong JG, Shi L, Liu J, Cao F, Ma YQ, Zhang Y. Predicting prostate cancer in men with PSA levels of 4-10 ng/mL: MRI-based radiomics can help junior radiologists improve the diagnostic performance. Sci Rep. 2023;13(1):4846. \u003c/li\u003e\n\u003cli\u003eLi M, Chen T, Zhao W, Wei C, Li X, Duan S, et al. Radiomics prediction model for the improved diagnosis of clinically significant prostate cancer on biparametric MRI. Quant Imaging Med Surg. 2020;10(2):368\u0026ndash;79. \u003c/li\u003e\n\u003cli\u003eCastillo T JM, Arif M, Starmans MPA, Niessen WJ, Bangma CH, Schoots IG, et al. Classification of Clinically Significant Prostate Cancer on Multi-Parametric MRI: A Validation Study Comparing Deep Learning and Radiomics. Cancers. 2021;14(1):12. \u003c/li\u003e\n\u003cli\u003ePenzkofer T, Tempany-Afdhal CM. Prostate Cancer Detection and Diagnosis: The Role of MR and its Comparison to other Diagnostic Modalities \u0026ndash; A Radiologist\u0026rsquo;s Perspective. NMR Biomed. 2014;27(1):10.1002/nbm.3002. \u003c/li\u003e\n\u003cli\u003eDuran R, Ronot M, Kerbaol A, Van Beers B, Vilgrain V. Hepatic hemangiomas: factors associated with T2 shine-through effect on diffusion-weighted MR sequences. Eur J Radiol. 2014;83(3):468\u0026ndash;78. \u003c/li\u003e\n\u003cli\u003eTeică RV, Șerbănescu MS, Florescu LM, Gheonea IA. Tumor Area Highlighting Using T2WI, ADC Map, and DWI Sequence Fusion on bpMRI Images for Better Prostate Cancer Diagnosis. Life. 2023;13(4):910. \u003c/li\u003e\n\u003cli\u003ePark SY, Kim CK, Park JJ, Park BK. Exponential apparent diffusion coefficient in evaluating prostate cancer at 3\u0026thinsp;T: preliminary experience. Br J Radiol. 2016;89(1058):20150470. \u003c/li\u003e\n\u003cli\u003eErickson BJ, Korfiatis P, Akkus Z, Kline TL. Machine Learning for Medical Imaging. Radiogr Rev Publ Radiol Soc N Am Inc. 2017;37(2):505\u0026ndash;15. \u003c/li\u003e\n\u003cli\u003eBreiman L. Random Forests. Mach Learn. 2001;45(1):5\u0026ndash;32. \u003c/li\u003e\n\u003cli\u003eKrauss W, Frey J, Heydorn Lagerl\u0026ouml;f J, Lid\u0026eacute;n M, Thunberg P. Radiomics from multisite MRI and clinical data to predict clinically significant prostate cancer. Acta Radiol Stockh Swed 1987. 2024;65(3):307\u0026ndash;17. \u003c/li\u003e\n\u003cli\u003eGaudiano C, Mottola M, Bianchi L, Corcioni B, Cattabriga A, Cocozza MA, et al. Beyond Multiparametric MRI and towards Radiomics to Detect Prostate Cancer: A Machine Learning Model to Predict Clinically Significant Lesions. Cancers. 2022;14(24):6156. \u003c/li\u003e\n\u003cli\u003eSultan MI, Huynh LM, Kamil S, Abdelaziz A, Hammad MA, Gin GE, et al. Utility of noninvasive biomarker testing and MRI to predict a prostate cancer diagnosis. Int Urol Nephrol. 2023;56(2):539. \u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1-3 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-genomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mgnm","sideBox":"Learn more about [BMC Medical Genomics](http://bmcmedgenomics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/mgnm/default.aspx","title":"BMC Medical Genomics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Prostate cancer, Radiomics, PRKY, DNA methylation, Urine specimen","lastPublishedDoi":"10.21203/rs.3.rs-6806278/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6806278/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e\u003cp\u003eTo enhance the diagnostic value of MRI for clinically significant prostate cancer (csPCa) and optimize the diagnostic process of prostate cancer (PCa), we developed a machine learning-based prediction model for csPCa combined with MRI features, clinical data, and PRKY promoter methylation level in urine samples.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThirty-nine patients who underwent prostate biopsy or transurethral laser enucleation of the prostate from 2022 to 2023 were selected for this study, and their clinical data and MRI images were obtained before the operation. The urine samples of these patients were collected after prostate massage. Methylation level of two PRKY promoter sites, cg05618150 and cg05163709, were tested through Methylation-Specific PCR. The PI-RADS score of each patient was estimated and the region of interest (ROI) was delineated. After being extracted by a plug-in of 3D-slicer, radiomics features were selected through LASSO regression and t-test. Selected radiomic features, methylation levels and clinical data were used for model construction through the random forest (RF) algorithm in Python. The model based on the PI-RADS score was also constructed for comparison with the radiomics model. The predictive efficiency of each model was analyzed by the area under the receiver operation characteristic (ROC) curve (AUC), and all the models have gone through 3-fold cross-validation.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eMethylation level of cg05163709 in csPCa patients was higher than that in clinically insignificant PCa and benign prostatic hyperplasia patients. The AUC of cg05163709 in csPCA prediction was 0.75. The AUC of the model combined with T2WI and ADC features was 0.91. And the model combined with radiomics features, Methylation level of cg05163709 and clinical data reached an AUC of 0.97, which was greater than that of the model based on the PI-RADS score (AUC\u0026thinsp;=\u0026thinsp;0.86).\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eAn effective prediction model for csPCa was successfully established by integrating T2WI features, clinical data, and methylation level of cg05163709 in urine specimens.\u003c/p\u003e","manuscriptTitle":"Constructing a Prediction Model for Clinically Significant Prostate Cancer Combined with Radiomics Features of MRI and PRKY Promoter Methylation Level in Urine Samples","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-23 08:52:26","doi":"10.21203/rs.3.rs-6806278/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2025-08-06T12:17:37+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"157159985349529984006603504980081504694","date":"2025-07-30T08:50:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-18T12:48:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-13T08:35:36+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-06-09T08:13:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-08T09:06:59+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Genomics","date":"2025-06-08T09:04:22+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-genomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mgnm","sideBox":"Learn more about [BMC Medical Genomics](http://bmcmedgenomics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/mgnm/default.aspx","title":"BMC Medical Genomics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d54d85cb-bc11-4a51-82b7-85028755a433","owner":[],"postedDate":"July 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-07-23T08:52:26+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-23 08:52:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6806278","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6806278","identity":"rs-6806278","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}