scholarly journals Prediction of prostate cancer by deep learning with multilayer artificial neural network

2018 ◽  
Vol 13 (5) ◽  
Author(s):  
Takumi Takeuchi ◽  
Mami Hattori-Kato ◽  
Yumiko Okuno ◽  
Satoshi Iwai ◽  
Koji Mikami
Keyword(s):  
Neural Network ◽  
Prostate Cancer ◽  
Artificial Neural Network ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Deep Learning ◽  
Logistic Regression Analysis ◽  
Prostate Biopsy ◽  
Stepwise Logistic Regression ◽  
Artificial Neural

Introduction: To predict the rate of prostate cancer detection on prostate biopsy more accurately, the performance of deep learning using a multilayer artificial neural network was investigated. Methods: A total of 334 patients who underwent multiparametric magnetic resonance imaging before ultrasonography guided transrectal 12-core prostate biopsy were enrolled in the analysis. Twenty-two non-selected variables, as well as selected ones by least absolute shrinkage and selection operator (Lasso) regression analysis and by stepwise logistic regression analysis, were input into the constructed multilayer artificial neural network (ANN) programs; 232 patients were used as training cases of ANN programs and the remaining 102 patients were for the test to output the probability of prostate cancer existence, accuracy of prostate cancer prediction, and area under the receiver operating characteristic (ROC) curve with the learned model. Results: With any prostate cancer objective variable, Lasso and stepwise regression analyses selected 12 and nine explanatory variables, respectively, from 22. Using trained ANNs with multiple hidden layers, the accuracy of predicting any prostate cancer in test samples was about 5–10% higher compared to that with logistic regression analysis (LR). The area under the curves (AUC) with multilayer ANN were significantly larger on inputting variables that were selected by the stepwise LR compared with the AUC with LR. The ANN had a higher net benefit than LR between prostate cancer probability cutoff values of 0.38 and 0.6. Conclusions: ANN accurately predicted prostate cancer without biopsy marginally better than LR. However, for clinical application, ANN performance may still need improvement.

scholarly journals Prediction of prostate cancer by deep learning with multilayer artificial neural network

2018 ◽  
Author(s):  
Takumi Takeuchi ◽  
Mami Hattori-Kato ◽  
Yumiko Okuno ◽  
Satoshi Iwai ◽  
Koji Mikami
Keyword(s):  
Neural Network ◽  
Prostate Cancer ◽  
Artificial Neural Network ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Deep Learning ◽  
Logistic Regression Analysis ◽  
Prostate Biopsy ◽  
Stepwise Logistic Regression ◽  
Artificial Neural

AbstractObjectivesTo predict the rate of prostate cancer detection on prostate biopsy more accurately, the performance of deep learning utilizing a multilayer artificial neural network was investigated.Materials and methodsA total of 334 patients who underwent multiparametric magnetic resonance imaging before ultrasonography-guided transrectal 12-core prostate biopsy were enrolled in the analysis. Twenty-two non-selected variables as well as selected ones by least absolute shrinkage and selection operator (Lasso) regression analysis and by stepwise logistic regression analysis were input into the constructed multilayer artificial neural network (ANN) programs. 232 patients were used as training cases of ANN programs, and the remaining 102 patients were for the test to output the probability of prostate cancer existence, accuracy of prostate cancer prediction, and area under the receiver operating characteristic (ROC) curve with the learned model.ResultsWith any prostate cancer objective variable, Lasso and stepwise regression analyses selected 12 and 9 explanatory variables from 22, respectively. In common between them, age at biopsy, findings on digital rectal examination, findings in the peripheral zone on MRI diffusion-weighted imaging, and body mass index were positively influential variables, while numbers of previous prostatic biopsy and prostate volume were negatively influential. Using trained ANNs with multiple hidden layers, the accuracy of predicting any prostate cancer in test samples was about 5-10% higher compared with that with logistic regression analysis (LR). The AUCs with multilayer ANN were significantly larger on inputting variables that were selected by the stepwise logistic regression compared with the AUC with LR. The ANN had a higher net-benefit than LR between prostate cancer probability cut-off values of 0.38 and 0.6.ConclusionANN accurately predicted prostate cancer without biopsy marginally better than LR. However, for clinical application, ANN performance may still need improvement.

scholarly journals Comparison of Artificial Neural Networks and Logistic Regression Analysis in Pregnancy Prediction Using the In Vitro Fertilization Treatment

2013 ◽  
Vol 35 (1) ◽  
pp. 39-48 ◽  
Author(s):  
Robert Milewski ◽  
Anna Justyna Milewska ◽  
Teresa Więsak ◽  
Allen Morgan
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Logistic Regression Analysis ◽  
Multivariable Logistic Regression Analysis ◽  
Vitro Fertilization ◽  
Multivariable Logistic Regression ◽  
Artificial Neural

Abstract Infertility is recognized as a major problem of modern society. Assisted Reproductive Technology (ART) is the one of many available treatment options to cure infertility. However, the efficiency of the ART treatment is still inadequate. Therefore, the procedure’s quality is constantly improving and there is a need to determine statistical predictors as well as contributing factors to the successful treatment. There is a concern over the application of adequate statistical analysis to clinical data: should classic statistical methods be used or would it be more appropriate to apply advanced data mining technologies? By comparing two statistical models, Multivariable Logistic Regression analysis and Artificial Neural Network it has been demonstrated that Multivariable Logistic Regression analysis is more suitable for theoretical interest but the Artificial Neural Network method is more useful in clinical prediction.

scholarly journals A nomogram to predict the upgrading rate of ISUP grades of RP in patients undergoing transrectal prostate biopsy and transperineal prostate biopsy

2021 ◽  
Author(s):  
Zhilei Zhang ◽  
Fei Qin ◽  
Guofeng Ma ◽  
Hang Yuan ◽  
Yongbo Yu ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Radical Prostatectomy ◽  
Logistic Regression Analysis ◽  
Prostate Biopsy ◽  
Curve Analysis ◽  
Decision Curve Analysis ◽  
Biopsy Tissue ◽  
Systematic Biopsy

Abstract Backgroud: This study was aimed to develop and internally validate a nomogram for risk of upgrade of ISUP (International Society of Urology Pathology) grade group from biopsy tissue to RP (radical prostatectomy) final histology.Methods: 166 patients with prostate cancer were retrospectively analyzed and divided into two groups based on ISUP upgrade status from needle biopsy to radical prostatectomy specimen, these being the 'ISUP upgrade' group and the 'no ISUP upgrade' group. Logistic regression analysis was used to predict the significant independent factors for ISUP upgrade. A nonogram was then developed based on these independent factors, which would predict risk of ISUP upgrade. The C-index, calibration plot, and decision curve analysis were used to assess the discrimination, calibration, and clinical usefulness of the predicting model. Internal validation was evaluated by using the bootstrapping validation. Results: There were 47 patients in the ISUP upgrade group and 119 patients in the no ISUP upgrade group respectively. Patients in the ISUP upgrade group tended to be of younger age, smaller PV (prostate volume), lower GS (Gleason score) of PB (prostate biopsy) tissue than the no ISUP upgrade group (p=0.043, p=0.041, p < 0.001, p =0.04, respectively). Multivariate logistic regression analysis showed that GS ≤6 (OR=14.236, P=0.001), prostate biopsy approach (TB-SB (transperineal prostate systematic biopsy) VS TR-SB (transrectal prostate systematic biopsy), OR=0.361, P=0.03) and number of positive cores < 10 (OR=0.396, P=0.04) were the independent risk factors for ISUP upgrade. A prediction nomogram model of ISUP upgrade was built based on these significant factors above, the area under the receiver operating characteristic (AUC) curve of which was 0.802. The C-index for the prediction nomogram was 0.798 (95%CI: 0.655–0.941) and the nomogram showed good calibration. High C-index value of 0.772 could still be reached in the interval validation. Decision curve analysis also demonstrated that the threshold value of RP-ISUP upgrade risk was 3% to 67%. Conclusion: A novel nomogram incorporating PSA, GS of PCa, ways of prostate biopsy and number of positive cores was built with a relatively good accuracy to assist clinicians to evaluate the risk of ISUP upgrade in the RP specimen, especially for the low-risk prostate cancer diagnosed by TR-SB.

scholarly journals The presence of chronic inflammation in positive prostate biopsy is associated with upgrading in radical prostatectomy

2021 ◽  
Vol 93 (3) ◽  
pp. 280-284
Author(s):  
Ekrem Guner ◽  
Yavuz Onur Danacioglu ◽  
Yusuf Arikan ◽  
Kamil Gokhan Seker ◽  
Salih Polat ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Radical Prostatectomy ◽  
Chronic Inflammation ◽  
Logistic Regression Analysis ◽  
Prostate Biopsy ◽  
Chronic Prostatitis ◽  
Significant Difference ◽  
The Impact

Objective: This study aimed to determine the predictive effect of the presence of chronic prostatitis associated with prostate cancer (PCa) in prostate biopsy on Gleason score upgrade (GSU) in radical prostatectomy (RP) specimens. Materials and methods: The data of 295 patients who underwent open or robotic RP with a diagnosis of localized PCa following biopsy were retrospectively analyzed. Patients were divided into two groups with and without GSU following RP. Predictive factors affecting GSU on biopsy were determined. The impact of chronic prostatitis associated with prostate cancer on GSU was examined via logistic regression analysis. Results: Out of 224 patients with Gleason 3+3 scores on biopsy, 145 (64.7%) had Gleason upgrade, and 79 (35.2%) had no upgrade. Whilst comparing the two groups with and without Gleason upgrade in terms of patient age, prostate-specific antigen (PSA) value, PSA density (PSAD), prostate volume (PV), neutrophil/lymphocyte (N/L) ratio, number of positive cores, percentage of positive cores, and Prostate Imaging Reporting and Data System version 2 score, no statistically significant difference was detected. The presence of chronic prostatitis associated with PCa was higher in the patient cohort with GSU in contrast to the other group (p < 0.001). According to the univariate logistic regression analysis, the presence of chronic prostatitis was identified to be an independent marker for GSU. Conclusions: Pathologists and urologists should be careful regarding the possibility of a more aggressive tumor in the presence of chronic inflammation associated with PCa because inflammation within PCa was revealed to be linked with GSU after RP.

Sign in / Sign up

Export Citation Format

Share Document