A stochastic model for prostate-specific antigen levels (Q1877147)

The authors investigate a stochastic model for a progression of the prostate-specific antigen (PSA) level. They consider PSA level as a continuous random variable governed by the stochastic equation \[ dX =[\alpha X(t)-ke^{-at}]dt +\sigma dZ(t) \] where \(Z(t)\) is a standard Wiener process, \(\sigma,\alpha, k\) and \(a\) are constants. Using a special change of variables \((t,x,p) \to ( \tau, y, Q)\) the authors reduce the forward Kolmogorov equation for the process \(X(t)\) distribution density \(p(t,x)\) to the heat equation for the function \(Q(\tau, y)\). This permits them to obtain the explicit expression for \(p(t,x)\), to investigate it and to give the implications for treatment.