A neurocomputational model for prostate carcinoma detection

Abstract

BACKGROUND Current guidelines for prostate carcinoma screening rely primarily on the digital rectal examination (DRE) and prostate specific antigen (PSA). Well described patient risk factors for prostate carcinoma also include age, ethnicity, family history, and complexed PSA. However, due to the nonlinear relation of each of these variables with prostate carcinoma, it is difficult to predict reliably each patient's risk based on linear univariate analysis. The authors investigated a neural network to model the risk of prostate carcinoma by seven readily available clinical features. METHODS The database for the current study comprised 3268 men recently evaluated for the early detection of prostate carcinoma. The seven clinical features evaluated included age, race, family history, International Prostate Symptom Score (IPSS), DRE, and total and complexed PSA. Three hundred forty‐eight subjects in the dataset included men with determined prostate biopsy outcomes and for whom at least 6 of 7 features were available. The dataset was divided randomly into a training set (60%) and a test set (40%), with n1/n2 cross‐validation used to evaluate model accuracy, and was modeled with linear and quadratic discriminant function analysis and a neural computational system. After a model with acceptable goodness of fit was achieved, reverse regression analysis using Wilks's generalized likelihood ratio test was performed to evaluate the statistical significance of each input variable. RESULTS The receiving operating characteristic (ROC) area for the neural computational system in the test set was 0.825, whereas total PSA and complexed PSA alone had ROC areas of 0.678 and 0.697, respectively. The ROC area of logistic regression in the test set was 0.510, linear discriminant function analysis was 0.674, and quadratic discriminant function analysis was 0.011. All were significantly less than the ROC area of the neural computational model (all Ps < 0.002). Reverse regression based on Wilks's generalized likelihood ratio test demonstrated each input feature to be highly significant to the model (all Ps ≪ 0.000001). CONCLUSIONS The authors modeled a combination of well described patient risk factors for prostate carcinoma using a neural computational system with acceptable goodness of fit. They demonstrated that each of the seven variates on which the model was based was critically significant to model performance. The authors presented this model for clinical use and suggested that clinicians use it in deciding to perform prostate biopsy. Cancer 2003. © 2003 American Cancer Society.

Is Prostate Specific Antigen Density more Important than in the Diagnosis of Prostate Cancer in Patients with a Total Prostate Specific Antigen Value <10 ng/mL?

“…Bu nedenle gereksiz biyopsilerden kaçınmak için çeşitli araştırmacı gruplar tarafından öngörü nomogramları geliştirilmiştir. Bu nomogramlar oluşturulurken, yaş, ırk, aile öyküsü, PRM, PSA, PSAD ve TRUS bulguları gibi değişkenler temel alınmaktadır (15,16). Çalışmamızdaki tüm hastaların PSA değerleri 2,5-10 ng/ ml arasındadır ve prostat kanseri oranımız %18,1 olarak saptanmıştır.…”

Internal validation of an artificial neural network for prostate biopsy outcome

“…The ratio of free PSA to total PSA (fPSA/tPSA or %fPSA) can increase specificity of tPSA 2 . Multivariate models combining tPSA, %fPSA or complexed PSA, age, status of digital rectal examination (DRE) or prostate volume are increasingly used to further improve PCa detection rate 3–8 . Comparison studies using receiver operator characteristic (ROC) curve analysis on different classification models like logistic regression (LR) based nomograms or artificial neural networks (ANN) have been recently reviewed 9–11 …”

“…2 Multivariate models combining tPSA, %fPSA or complexed PSA, age, status of digital rectal examination (DRE) or prostate volume are increasingly used to further improve PCa detection rate. [3][4][5][6][7][8] Comparison studies using receiver operator characteristic (ROC) curve analysis on different classification models like logistic regression (LR) based nomograms or artificial neural networks (ANN) have been recently reviewed. [9][10][11] With the increasing use of free available internet-based models for PCa detection as provided by Finne et al (two models at http:// urologie.charite.de Link: ProstataClass), one must pay more attention to validation studies of these models.…”

Clinical utility of human glandular kallikrein 2 within a neural network for prostate cancer detection

“…However, logistic regression has a limited capacity to handle very complex data, and ANNs are better for detection, especially of nonlinear relationships among multiple variables [25]. These and other ANNs were trained with different input variables such as tPSA, f/tPSA, complexed PSA, age, race, family history, prostate volume, prostate volume indexes, findings from TRUS, or status of DRE [21–24,26–30]. As prostate volume (measured by TRUS) and status of DRE are more subjective measurements, they were not included in the present study.…”