Application Value of Radiomic Nomogram in the Differential Diagnosis of Prostate Cancer and Hyperplasia

Shaogao Gui; Min Lan; Chaoxiong Wang; Si Nie; Bing Fan

doi:10.3389/fonc.2022.859625

Application Value of Radiomic Nomogram in the Differential Diagnosis of Prostate Cancer and Hyperplasia

Front Oncol. 2022 Apr 14:12:859625. doi: 10.3389/fonc.2022.859625. eCollection 2022.

Authors

Shaogao Gui¹, Min Lan², Chaoxiong Wang¹, Si Nie¹, Bing Fan¹

Affiliations

¹ Department of Radiology, Jiangxi Provincial People's Hospital, The First Affiliated Hospital of Nanchang Medical College, Nanchang, China.
² Department of Orthopedics, Jiangxi Provincial People's Hospital, The First Affiliated Hospital of Nanchang Medical College, Nanchang, China.

Abstract

Objective: Prostate cancer and hyperplasia require different treatment strategies and have completely different outcomes; thus, preoperative identification of prostate cancer and hyperplasia is very important. The purpose of this study was to evaluate the application value of magnetic resonance imaging (MRI)-derived radiomic nomogram based on T2-weighted images (T2WI) in differentiating prostate cancer and hyperplasia.

Materials and methods: One hundred forty-six patients (66 cases of prostate cancer and 80 cases of prostate hyperplasia) who were confirmed by surgical pathology between September 2019 and September 2019 were selected. We manually delineated T2WI of all patients using ITK-SNAP software and radiomic analysis using Analysis Kit (AK) software. A total of 396 tumor texture features were extracted. Subsequently, the effective features were selected using the LASSO algorithm, and the radiomic feature model was constructed. Next, combined with independent clinical risk factors, a multivariate Logistic regression model was used to establish a radiomic nomogram. The receiver operator characteristic (ROC) curve was used to evaluate the prediction performance of the radiomic nomogram. Finally, the clinical application value of the nomogram was evaluated by decision curve analysis.

Results: The PSA and the selected imaging features were significantly correlated with the differential diagnosis of prostate cancer and hyperplasia. The radiomic model had good discrimination efficiency for prostate cancer and hyperplasia. The training set (AUC = 0.85; 95% CI: 0.77-0.92) and testing set (AUC = 0.84; 95% CI: 0.72-0.96) were effective. The radiomic nomogram, combined with the radiomic characteristics of MRI and independent clinical risk factors, showed better differentiation efficiency in the training set (AUC = 0.91; 95% CI: 0.85-0.97) and testing set (AUC = 0.90; 95% CI: 0.81-0.99). The decision curve showed the clinical application value of the radiomic nomogram.

Conclusion: The radiomic nomogram of T2-MRI combined with clinical risk factors can easily identify prostate cancer and hyperplasia. It also provides suggestions for further clinical events.

Keywords: hyperplasia; magnetic resonance imaging; prostate cancer; radiomic; textural features.