A fuzzy logic based-method for prognostic decision making in breast and prostate cancers

Area of Science:

• Oncology
• Biostatistics
• Machine Learning

Background:

• Accurate oncological prognosis is crucial for effective patient management and treatment planning.
• The proliferation of novel prognostic markers necessitates advanced methods for identifying clinically significant ones.
• Complex, nonlinear interactions among markers pose challenges for traditional prognostic models.

Purpose of the Study:

• To evaluate the fuzzy k-nearest neighbor (FK-NN) classifier for oncological prognosis.
• To assess the FK-NN method's ability to provide a certainty degree for prognostic decisions.
• To compare FK-NN with logistic regression and artificial neural networks for prognostic marker assessment.

Main Methods:

• Utilized the fuzzy k-nearest neighbor (FK-NN) classifier.
• Compared FK-NN with logistic regression (statistical method).
• Compared FK-NN with multilayer feedforward backpropagation neural networks (artificial neural network).
• Employed breast and prostate cancer datasets as benchmarks.

Main Results:

• The FK-NN-based method achieved the highest predictive accuracy.
• FK-NN produced a more reliable prognostic marker model.
• FK-NN outperformed both statistical and artificial neural network methods in accuracy and reliability.

Conclusions:

• The FK-NN classifier is a highly accurate and reliable tool for oncological prognosis.
• Fuzzy logic offers advantages in assessing prognostic markers with complex interactions.
• FK-NN provides a robust alternative to conventional statistical and neural network approaches in cancer prognostics.