Prediction of prostate cancer by the general discriminant analysis models

Cover Page

Abstract


The study presents a retrospective analysis of risk factors for prostate cancer in 122 men. First, a comparative analysis of many factors: age, weight, bad habits, levels of testosterone and of prostate-specific antigen (PSA), comorbidities, blood group, RH factor, manual rectal investigation, the results of the prostate ultrasound was conducted. Next, using general models of discriminant analysis, five of the most significant risk factors for prostate cancer were identified: inflammatory diseases of the kidneys and bladder, increased PSA blood test, blood group, overweight patient. Applied discriminant equation containing the five most informative features, allows up to 97.5% to predict the likelihood of developing prostate cancer.

A A Kostin

Peoples' Friendship University of Russia

Email: info@mnioi.ru
Department of urology, oncology and radiology Medical Institute; P.A. Herzen Moscow Oncology Research Institute

A T Asratov

Peoples' Friendship University of Russia

Email: dr.asad1982@yandex.ru
Department of urology, oncology and radiology Medical Institute

N G Kulchenko

Peoples' Friendship University of Russia; P.A. Herzen Moscow Oncology Research Institute

Email: kle-kni@mail.ru
Department of histology, cytology, embryology Medical Institute

A O Tolkachev

P.A. Herzen Moscow Oncology Research Institute

Email: alext06@yandex.ru

Views

Abstract - 77

PDF (Russian) - 41


Copyright (c) 2015 Костин А.А., Асратов А.Т., Кульченко Н.Г., Толкачев А.О.

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.