MP74-16 AVOIDING UNNECESSARY PROSTATE BIOPSY TO DETECT PRIMARY PROSTATE CANCER: THE NON-INVASIVE DIAGNOSIS VALUE OF 68 GA-PSMA PET/CT-BASED RADIOMICS ANALYSIS WITH MACHINE LEARNING
Jiaxian Chen,Yongxiang Tang,Lin Qi,Xiaomei Gao,Shuo Hu,Yi Cai
DOI: https://doi.org/10.1097/01.ju.0001008632.59099.b9.16
2024-01-01
Abstract:You have accessJournal of UrologyProstate Cancer: Detection & Screening VI (MP74), Moderated Poster 741 May 2024MP74-16 AVOIDING UNNECESSARY PROSTATE BIOPSY TO DETECT PRIMARY PROSTATE CANCER: THE NON-INVASIVE DIAGNOSIS VALUE OF 68GA-PSMA PET/CT-BASED RADIOMICS ANALYSIS WITH MACHINE LEARNING Jiaxian Chen, Yongxiang Tang, Lin Qi, Xiaomei Gao, Shuo Hu, and Yi Cai Jiaxian ChenJiaxian Chen , Yongxiang TangYongxiang Tang , Lin QiLin Qi , Xiaomei GaoXiaomei Gao , Shuo HuShuo Hu , and Yi CaiYi Cai View All Author Informationhttps://doi.org/10.1097/01.JU.0001008632.59099.b9.16AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Multiparametric magnetic resonance imaging (mpMRI) is the most commonly used imaging to screening prostate cancer(PCa). However, with the intervention of mpMRI, prostate biopsies still tend to dectect benign prostate diseases (BPD) in half of the patients. A biopsy-free diagnostic pathway in PCa is limited by the low specificity of mpMRI. Compared with mpMRI, 68Ga-prostate specific membrane antigen (68Ga-PSMA) positron emission tomography/computed tomography (PET/CT) offers greater specificity in detecting PCa lesions. Recent studies have showed that 68Ga-PSMA PET/CT has the potential to open up a non-invasive diagnostic pathway in PCa. However, the relatively small number of enrolled patients in these studies(especially the lack of BPD patients) may limit the statistical validation and cause bias to some extent. Besides, the performance of 68Ga-PSMA PET/CT radiomics analysis in detecting PCa has not been widely evaluated. Hence, the purpose of this study was to develop radiomics models based on 68Ga -PSMA PET/CT and evaluate its diagnostic value for the detection of primary PCa in a relatively larger cohort. METHODS: This retrospective study included consecutive patients with suspected PCa who underwent systematic biopsy (SB) and PSMA PET-targeted biopsy (TB) after 68Ga-PSMA PET/CT. The whole prostate gland was adopted as volume of interest(VOI). Classic minimum redundancy maximum relevance (mRMR) algorithm and least absolute shrinkage and selection operator (LASSO) algorithm were used to select significant distinguishable features. 8 separate machine learning(ML) classifiers namely eXtreme Gradient Boosting (XGBoost), Decision tree, Naïve Bayes, Neural Network(NNET), K-nearest neighbor (KNN), Random forest, Logistic regression, Support vector machine (SVM) were trained and tested with ten-fold cross-validation in each analysis subgroup. The sensitivity, specificity, accuracy, positive predictive value(PPV), negative predictive value(NPV), and area under the receiver operating characteristic curve (AUC) were calculated for each model. RESULTS: 349 patients were eniligible for this subgroup analysis including 97 patients with BPD, 20 patients with PCa classified into the GS=3+3(ciPCa) and 232 patients with PCa classified into the GS ≥3+4(csPCa). For PCa prediction, the radiomics model based on Random forest algorithm achieved a specificity and PPV of 100%. The Logistic regression algorithm achieved the highest AUC of 0.924 (with sensitivity, 85.3%; specificity, 93.1%; accuracy, 87.5% ; PPV, 97%; NPV, 71.1% ). For csPCa prediction, the Decision tree algorithm achieved the highest AUC of 0.921 (with sensitivity, 92.6%; specificity, 82.9%; accuracy, 88.3%; PPV, 91.3%; NPV, 0.853%). CONCLUSIONS: To the best of our knowledge, the present study included the largest number of patients(97 individuals in BPD cohort, 252 individuals in PCa cohort, respectively), providing a non-invasive and quantitative method for predicting PCa, which could be used to gauge PCa risk prior to unnecessary biopsies. Source of Funding: This research was supported by National Natural Science Foundation of China (82272907,81974397), the key Research and Development program of Hunan Province(2021SK2014), National Key Research and Development Plan (2017YFC0908004), Clinical Big Data System Construction Project Fund of Xiangya Hospital (No. 33020125030) and the Xiangya Famous Doctor Fund of Central South University (33020123007) © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e1199 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Jiaxian Chen More articles by this author Yongxiang Tang More articles by this author Lin Qi More articles by this author Xiaomei Gao More articles by this author Shuo Hu More articles by this author Yi Cai More articles by this author Expand All Advertisement PDF downloadLoading ...