Building Biomarker Combinations for Korean Ovarian Cancer Screening Using Statistics and Machine Learning

Abstract

Early screening using appropriate biomarkers is helpful for the effective treatment of ovarian cancer. CA-125, the most widely used biomarker for the diagnosis of ovarian cancer, has high false positive and false negative rates. We introduce an approach for determining an appropriate combination of biomarkers known to be highly related to ovarian cancer among 21 predetermined biomarkers. Sera representing 27 cases and 31 controls from women undergoing surgery were examined using high-throughput, multiplexed bead-based immunoassays. Student’s t-test and a genetic algorithm (GA) were employed and compared for building the proper combination of two to four biomarkers. The combinations selected by both methods were compared with a 5-fold cross validation of the LDA classifier. The combination of four markers chosen by the GA had the best performance in regards to accuracy, with sensitivity and specificity of 81% and 100%, respectively.

Systematic Analysis of Ovarian Cancer Empowered with Machine and Deep Learning: A Taxonomy and Future Challenges

“…The five-fold cross-validation performed on the 21 biomarkers for better specificity and sensitivity of dataset. The 0.92% accuracy achieved on with the combination of all biomarkers [73]. Gene selection in large micro-array data needed for the classification of OC, for a purpose Scatter search is implies on gene selection for ML algorithms-support vector machine and decision tree.…”

Model‐Independent Evaluation of Tumor Markers and a Logistic‐Tree Approach to Diagnostic Decision Support

“…Their prediction procedures seem to be "rough" in combining multiple independent variables and are not suited for accommodating the contradictory effect of sensitivity and specificity. By this rationale, several models originated from machine learning such as support vector machine (SVM) [4], classification tree [9], evolutionary algorithm [10][11][12] and neural networks [9,[13][14][15][16][17] have been widely applied for both evaluation of diagnostic values and diagnostic decision support. Sun et al [4] applied SVM on a sample set with 12 tumor markers, and it was found that compared with parallel test, SVM achieved more specificity increase at the expense of less sensitivity decrease.…”