Portal de Periódicos da CAPES

MP31-03 DECODING CLINICALLY SIGNIFICANT PROSTATE CANCER IN BIOPSY NAIVE PATIENTS: A MACHINE LEARNING APPROACH FROM A TERTIARY REFERRAL CENTER

2024; Lippincott Williams & Wilkins; Volume: 211; Issue: 5S Linguagem: Inglês

10.1097/01.ju.0001008936.35187.0b.03

1527-3792

Cesare Saitta, Filippo Dagnino, Vittorio Fasulo, Pierpaolo Avolio, Marco Paciotti, Nicola Frego, Davide Maffei, Paola Arena, Edoardo Beatrici, Fabio Decarne, Paolo Casale, Massimo Lazzeri, Alberto Saita, Rodolfo Hurle, Giorgio Guazzoni, Nicolò Maria Buffi, Giovanni Lughezzani,

Radiomics and Machine Learning in Medical Imaging

You have accessJournal of UrologyProstate Cancer: Detection & Screening III (MP31)1 May 2024MP31-03 DECODING CLINICALLY SIGNIFICANT PROSTATE CANCER IN BIOPSY NAIVE PATIENTS: A MACHINE LEARNING APPROACH FROM A TERTIARY REFERRAL CENTER Cesare Saitta, Filippo Dagnino, Vittorio Fasulo, Pierpaolo Avolio, Marco Paciotti, Nicola Frego, Davide Maffei, Paola Arena, Edoardo Beatrici, Fabio Decarne, Paolo Casale, Massimo Lazzeri, Alberto Saita, Rodolfo Hurle, Giorgio Guazzoni, Nicolò Buffi, and Giovanni Lughezzani Cesare SaittaCesare Saitta , Filippo DagninoFilippo Dagnino , Vittorio FasuloVittorio Fasulo , Pierpaolo AvolioPierpaolo Avolio , Marco PaciottiMarco Paciotti , Nicola FregoNicola Frego , Davide MaffeiDavide Maffei , Paola ArenaPaola Arena , Edoardo BeatriciEdoardo Beatrici , Fabio DecarneFabio Decarne , Paolo CasalePaolo Casale , Massimo LazzeriMassimo Lazzeri , Alberto SaitaAlberto Saita , Rodolfo HurleRodolfo Hurle , Giorgio GuazzoniGiorgio Guazzoni , Nicolò BuffiNicolò Buffi , and Giovanni LughezzaniGiovanni Lughezzani View All Author Informationhttps://doi.org/10.1097/01.JU.0001008936.35187.0b.03AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Detection of clinically significant Prostate Cancer (csPca) is a quality-of-care concern. While multiparametric MRI has been linked to improved detection of csPca, micro-Ultrasound (micro-US) has demonstrated equivalent effectiveness in detecting csPca when compared to MRI. Herein we sought to create a machine learning (ML) model capable of predicting the probability of csPca, which may assist in clinical decision making. METHODS: We queried our prospective database for biopsy naïve patients who underwent biopsy at our center. Micro-US sonography was performed during the biopsy. Type of biopsy performed (target biopsy only vs. target+systematic/or only systematic) was dictated by surgeons expertise and based on imaging reports. CsPca was defined as any Gleason group grade>1. Primary outcome was the development of a preoperative model capable of predict csPca. Secondary outcome was to juxtapose ML performances to traditional logistic regression (LR) model. ML algorithm evaluated were random forest, extreme gradient boosting (XGBoost), and light gradient boosting machine (LightGBM). Model's performances were evaluated with receiver operator characteristic curve (ROC) analysis using estimated area under the curve (AUC), accuracy, and F1 score. RESULTS: Overall, 848 patients were analyzed. LR model using PSA density>0.15, suspect lesion detected at micro-US and MRI (PRIMUS and PIRADS≥3), increasing age, positive digital rectal examination, target biopsy only vs. target biopsy + systematic demonstrated: Accuracy: 0.71, F1 Score: 0.64, AUC: 0.77. Every ML model demonstrated superior performances when compared to LR (Figure 1). Among ML model Radom Forest demonstrated the best performances (Accuracy: 0.86, F1 Score: 0.84, AUC: 0.95). CONCLUSIONS: Combining clinical features, serum biomarkers and imaging findings, we have developed a point of care model capable of accurartely predicting the presence of csPca. These findings may refine clinical decision making with respect to identification of patients with increased likelihood of csPca and with respect for the type of biopsy template performed. External validation is requisite. Download PPT Source of Funding: None © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e504 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Cesare Saitta More articles by this author Filippo Dagnino More articles by this author Vittorio Fasulo More articles by this author Pierpaolo Avolio More articles by this author Marco Paciotti More articles by this author Nicola Frego More articles by this author Davide Maffei More articles by this author Paola Arena More articles by this author Edoardo Beatrici More articles by this author Fabio Decarne More articles by this author Paolo Casale More articles by this author Massimo Lazzeri More articles by this author Alberto Saita More articles by this author Rodolfo Hurle More articles by this author Giorgio Guazzoni More articles by this author Nicolò Buffi More articles by this author Giovanni Lughezzani More articles by this author Expand All Advertisement PDF downloadLoading ...