Principles of Interpolating Prognostication in Oncology

Abstract

Abstract
The precise prediction in oncology plays a critical role in prevention, treatment, assessment of the effectiveness of methods of treatment and treatment outcomes for cancer patients. Some medical researches claim that statistical processing (factor analysis) permits to precisely individualize the prognosis, which determines the possibility of an individual approach to monitoring and postoperative treatment of patients. However, factor analysis permits only creating a matrix of factors that can significantly describe the study area, but it makes no sense to talk about factor prognostication, and even more about an error-free individualized prognosis, which determines the possibility of an individual approach to monitoring and postoperative treatment of patients. Even mathematical calculation of the existing pairwise correlation between all the studied symptoms (correlation matrix) and, in some cases, determination (predicting) with a certain accuracy of the relationship (regression) between a pair of studied symptoms or an investigated symptom with several symptoms using a polynomial regression program, or multiple linear regression and stepwise multiple regression programs do not permit to make the precise prognosis. The most precise prognostication method in computational mathematics today is the interpolating method. In the paper, along with the description of existing interpolating methods, the two developed principles of interpolating prognostication in oncology are presented: the minimizing of the difference (error) between the known and calculated by different polynomial interpolating methods numerical values of the investigated symptom; the selection as the values of the investigated symptom the maximal values of statistical frequencies of the observed patients in the given intervals of conducting study. These principles led to the development of an algorithm and then a system of computer programs for optimal interpolating prognostication of the investigated symptoms in oncology. Application of the developed system in clinical conditions permitted to identify a splash ("hump") of patients with metastases after surgery and to determine the effect of using of various methods of preventive treatment after surgery on the increase of the time before the appearance of metastases among the patients, i.e. to determine the effectiveness of the applied treatments.

Keywords: Computational mathematics, interpolation, optimal interpolating prognostication, the effectiveness of the applied treatments of malignant new formations, precise prognosis of time of metastases appearance after surgery

Cite this Article

M. Shoikhedbrod. Principles of Interpolating Prognostication in Oncology. Research & Reviews: Journal of Oncology and Hematology. 2020; 9(2): 39–51p