To investigate whether magnetic resonance (MR) radiomic features combined with machine learning may aid in predicting extraprostatic extension (EPE) in high- and non-favorable intermediate-risk patients with prostate cancer.
Purpose
To investigate the diagnostic performance of radiomics to detect EPE.
Material and Methods
MR radiomic features were extracted from 228 patients, of whom 86 were diagnosed with EPE, using prostate and lesion segmentations. Prediction models were built using Random Forest. Further, EPE was also predicted using a clinical nomogram and routine radiological interpretation and diagnostic performance was assessed for individual and combined models.
Results
The MR radiomic model with features extracted from the manually delineated lesions performed best among the radiomic models with an area under the curve (AUC) of 0.74. Radiology interpretation yielded an AUC of 0.75 and the clinical nomogram (MSKCC) an AUC of 0.67. A combination of the three prediction models gave the highest AUC of 0.79.
Conclusion
Radiomic analysis combined with radiology interpretation aid the MSKCC nomogram in predicting EPE in high- and non-favorable intermediate-risk patients.
MottetNBellmuntJBollaM, et al.EAU-ESTRO-SIOG Guidelines on Prostate Cancer. Part 1: screening, diagnosis, and local treatment with curative intent. Eur Urol2017;
71:618–629.
2.
KattanMWWheelerTMScardinoPT.Postoperative nomogram for disease recurrence after radical prostatectomy for prostate cancer. J Clin Oncol1999;
17:1499–1507.
3.
EpsteinJICarmichaelMJPizovG, et al.
Influence of capsular penetration on progression following radical prostatectomy: a study of 196 cases with long-term followup. J Urol1993;
150:135–141.
4.
BallMWPartinAWEpsteinJI.Extent of extraprostatic extension independently influences biochemical recurrence-free survival: evidence for further pT3 subclassification. Urology2015;
85:161–164.
5.
DannemanDWiklundFWiklundNP, et al.
Prognostic significance of histopathological features of extraprostatic extension of prostate cancer. Histopathology2013;
63:580–589.
6.
SteuberTGraefenMHaeseA, et al.
Validation of a nomogram for prediction of side specific extracapsular extension at radical prostatectomy. J Urol2006;
175:939–944.
7.
Di BlasioCJRheeACChoD, et al.
Predicting clinical end points: treatment nomograms in prostate cancer. Semin Oncol2003;
30:567–586.
8.
BrigantiAChunFK-HSaloniaA, et al.
Validation of a nomogram predicting the probability of lymph node invasion among patients undergoing radical prostatectomy and an extended pelvic lymphadenectomy. Eur Urol2006;
49:1017–1019.
9.
EiflerJBFengZLinBM, et al.
An updated prostate cancer staging nomogram (Partin tables) based on cases from 2006 to 2011. BJU Int2013;
111:22–29.
10.
de RooijMHamoenEHJWitjesJA, et al.
Accuracy of magnetic resonance imaging for local staging of prostate cancer: a diagnostic meta-analysis. Eur Urol2016;
70:233–245.
11.
FüttererJJBrigantiADe VisschereP, et al.
Can clinically significant prostate cancer be detected with multiparametric magnetic resonance imaging? A systematic review of the literature. Eur Urol2015;
68:1045–1053.
12.
WangJWuC-JBaoM-L, et al.
Using support vector machine analysis to assess PartinMR: A new prediction model for organ-confined prostate cancer. J Magn Reson Imaging2018;
48:499–506.
13.
SomfordDMHamoenEHFuttererJJ, et al.
The predictive value of endorectal 3 Tesla multiparametric magnetic resonance imaging for extraprostatic extension in patients with low, intermediate and high risk prostate cancer. J Urol2013;
190:1728–1734.
14.
BjurlinMARosenkrantzABSarkarS, et al.
Prediction of prostate cancer risk among men undergoing combined MRI-targeted and systematic biopsy using novel pre-biopsy nomograms that incorporate mri findings. Urology2018;
112:112–120.
15.
GuptaRTBrownAFSilvermanRK, et al.
Can radiologic staging with multiparametric MRI enhance the accuracy of the partin tables in predicting organ-confined prostate cancer?AJR Am J Roentgenol2016;
207:87–95.
16.
FariaRSoaresMOSpackmanE, et al.
Optimising the diagnosis of prostate cancer in the era of multiparametric magnetic resonance imaging: a cost-effectiveness analysis based on the Prostate MR Imaging Study (PROMIS). Eur Urol2018;
73:23–30.
17.
ReisæterLARFüttererJJLosnegårdA, et al.
Optimising preoperative risk stratification tools for prostate cancer using mpMRI. Eur Radiol2018;
28:1016–1026.
18.
MaubonTBrangerNBastideC, et al.
Impact of the extent of extraprostatic extension defined by Epstein’s method in patients with negative surgical margins and negative lymph node invasion. Prostate Cancer Prostatic Dis2016;
19:317–321.
19.
LeeHKimCKParkBK, et al.
Accuracy of preoperative multiparametric magnetic resonance imaging for prediction of unfavorable pathology in patients with localized prostate cancer undergoing radical prostatectomy. World J Urol2017;
35:929–934.
20.
YipSSFAertsHJWL.Applications and limitations of radiomics. Phys Med Biol2016;
61:R150.
21.
KumarVGuYBasuS, et al.
Radiomics: the process and the challenges.Magn Reson Imaging2012;
30:1234–1248.
22.
ParmarCGrossmannPBussinkJ, et al.
Machine learning methods for quantitative radiomic biomarkers.Sci Rep2015;
5:13087.
23.
MadabhushiAFeldmanMDMetaxasDN, et al.
Automated detection of prostatic adenocarcinoma from high-resolution ex vivo MRI. IEEE Trans Med Imaging2005;
24:1611–1625.
24.
AlgoharyAViswanathSShiradkarR, et al.
Radiomic features on MRI enable risk categorization of prostate cancer patients on active surveillance: Preliminary findings. J Magn Reson Imaging2018. Doi: 10.1002/jmri.25983.
25.
WangJWuC-JBaoM-L, et al.
Machine learning-based analysis of MR radiomics can help to improve the diagnostic performance of PI-RADS v2 in clinically relevant prostate cancer. Eur Radiol2017;
27:4082–4090.
26.
GnepKFargeasAGutierrez-CarvajalRE, et al.
Haralick textural features on T2 -weighted MRI are associated with biochemical recurrence following radiotherapy for peripheral zone prostate cancer. J Magn Reson Imaging2017;
45:103–117.
27.
VallieresMFreemanCRSkameneSR, et al.
A radiomics model from joint FDG-PET and MRI texture features for the prediction of lung metastases in soft-tissue sarcomas of the extremities. Phys Med Biol2015;
60:5471–5496.
28.
GinsburgSBAlgoharyAPahwaS, et al.
Radiomic features for prostate cancer detection on MRI differ between the transition and peripheral zones: Preliminary findings from a multi-institutional study. J Magn Reson Imaging2017;
46:184–193.
29.
KhalvatiFWongAHaiderMA.Automated prostate cancer detection via comprehensive multi-parametric magnetic resonance imaging texture feature models. BMC Med Imaging2015;
15:27.
30.
CameronAKhalvatiFHaiderMA, et al.
MAPS: a quantitative radiomics approach for prostate cancer detection. IEEE Trans Biomed Eng2016;
63:1145–1156.
31.
LitjensGJSElliottRShihNNC, et al.
Computer-extracted features can distinguish noncancerous confounding disease from prostatic adenocarcinoma at multiparametric MR imaging.Radiology2016;
278:135–145.
32.
ShiradkarRGhoseSJamborI, et al.
Radiomic features from pretreatment biparametric MRI predict prostate cancer biochemical recurrence: Preliminary findings. J Magn Reson Imaging2018;
48:1626–1636.
33.
WeinrebJCBarentszJOChoykePL, et al.
PI-RADS Prostate Imaging – Reporting and Data System: 2015, Version 2. Eur Urol2016;
69:16–40.
JenkinsonMBeckmannCFBehrensTEJ, et al.
FSL. NeuroImage2012;
62:782–790.
36.
Magi-GalluzziCEvansAJDelahuntB, et al.
International Society of Urological Pathology (ISUP) Consensus Conference on Handling and Staging of Radical Prostatectomy Specimens. Working group 3: extraprostatic extension, lymphovascular invasion and locally advanced disease. Mod Pathol2011;
24:26–38.
37.
MehralivandSShihJHHarmonS, et al.
A grading system for the assessment of risk of extraprostatic extension of prostate cancer at multiparametric MRI. Radiology2019;
290:709–719.
38.
YushkevichPAPivenJHazlettHC, et al.
User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. NeuroImage2006;
31:1116–1128.
39.
AvantsBBTustisonNJSongG, et al.
A reproducible evaluation of ANTs similarity metric performance in brain image registration. NeuroImage2011;
54:2033–2044.
40.
JurekJKocińskiMMaterkaA, et al.
Rule-based data-driven approach for computer aided diagnosis of the peripheral zone prostate cancer from multiparametric MRI: Proof of concept. In: 2017 Signal Processing:Algorithms, Architectures, Arrangements, and Applications (SPA), Poznan, Poland, 2017 September 20--22; pp. 90--95. New Jersey, US: IEEE.
41.
BreimanL.Random Forests. Mach Learn2001;
45:5–32.
42.
HoTK. Random decision forests. In: Proceedings of the Third International Conference on Document Analysis and Recognition Vol. 1, Montreal, Canada, 1995 August 14--16; New Jersey, US: IEEE.
43.
PedregosaFVaroquauxGGramfortA, et al.
Scikit-learn: machine learning in {P}ython. J Mach Learn Res2011;
12:2825–2830.
44.
DeLongERDeLongDMClarke-PearsonDL.Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics1988;
44:837–845.
45.
OhoriMKattanMWKohH, et al.
Predicting the presence and side of extracapsular extension: a nomogram for staging prostate cancer. J Urol2018;
171:1844–1849.
46.
DuvnjakPSchulmanAAHoltzJN, et al.
Multiparametric Prostate MR imaging: impact on clinical staging and decision making. Radiol Clin North Am2018;
56:239–250.
47.
FengTSSharif-AfsharARWuJ, et al.
Multiparametric MRI improves accuracy of clinical nomograms for predicting extracapsular extension of prostate cancer. Urology2018;
86:332–337.
48.
WeaverJKKimEHVetterJM, et al.
Prostate magnetic resonance imaging provides limited incremental value over the Memorial Sloan Kettering Cancer Center preradical prostatectomy nomogram. Urology2018;
113:119–128.
49.
GuoYGaoYShenD.Deformable MR prostate segmentation via deep feature learning and sparse patch matching. IEEE Trans Med Imaging2016;
35:1077–1089.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
