Prognostic Projection of Breast Cancer: Artificial Intelligence Modelling

There is a dearth of breast cancer data in developing
countries, especially Nigeria. This has made it difficult
to make satisfactory population based inference with
regards to prognosis of the disease. While it is known
that the prognosis of breast cancer is generally bad and indeed worse among African women diagnosed, it is not known to what extent and what factors are more
contributory than others. With unavailable data, it is
difficult to use classical mathematical tools to make
prognostic projections. An artificial intelligence
modeling system – the fuzzy inference system – is
adequate for data analysis in this situation.
The traditional factors that predict recurrence of the
disease or death have been analyzed in many studies.
This study analyzed four input variables (Age at diagnosis, Age, Stage, and Race) and one output
variable (five year survival rate). Data used was from
the National Cancer Institute. The rate of risk of attainment of each input variable expresses the output variable. A database was established. The prognosis
(five year survival rate) of breast cancer can be made by varying the input data. This calculated result takes
into account all uncertainties and inaccuracies related to the nature of the input variables.