The Application of Data Mining Techniques to Oral Cancer Prognosis

Abstract

This study adopted an integrated procedure that combines the clustering and classification features of data mining technology to determine the differences between the symptoms shown in past cases where patients died from or survived oral cancer. Two data mining tools, namely decision tree and artificial neural network, were used to analyze the historical cases of oral cancer, and their performance was compared with that of logistic regression, the popular statistical analysis tool. Both decision tree and artificial neural network models showed superiority to the traditional statistical model. However, as to clinician, the trees created by the decision tree models are relatively easier to interpret compared to that of the artificial neural network models. Cluster analysis also discovers that those stage 4 patients whose also possess the following four characteristics are having an extremely low survival rate: pN is N2b, level of RLNM is level I-III, AJCC-T is T4, and cells mutate situation (G) is moderate.

Construction of a flow chart–like risk prediction model of ganciclovir‐induced neutropaenia including severity grade: A data mining approach using decision tree

“…The predictive accuracies were the same (84.6%), and the misclassification risk of the DT model estimated by 10‐fold cross‐validation was 15.4% ± 1.8%. Although the cut‐off values are not well established, previous reports have determined favourable accuracies of approximately 80%‐90% and misclassification risk of 10%‐20% . This suggests that our results can be considered favourable.…”

“…By following this flow chart–like model, we could estimate the risk of neutropaenia easily and quantitatively. Importantly, DT models do not require special statistical analysis software, unlike other data mining methods, such as neural networks and support vector machines, which facilitates the interpretation and application of the DT model built in this study to clinical practice. Our risk prediction model can be applied to the following situations: first, for the management of high‐risk groups such as those having baseline ANC <3854 cells/mm 3 and duration of therapy ≥15 days, indicative of possible neutropaenia and/or leading to recommendations for alternative antiviral agents.…”

“…Decision tree (DT) analysis is a conventional technique of data mining consisting of a flow chart–like framework that shows the different outcomes for a series of decisions . By following the flow chart, users can estimate combinations of risk factors that may increase the probability of certain events . Recently, several reports encompassing DT analysis demonstrated its usefulness in the medical field, for example, for defining prognostic factors in various diseases .…”

“…10,11 Recently, several reports encompassing DT analysis demonstrated its usefulness in the medical field, for example, for defining prognostic factors in various diseases. [11][12][13] In our previous research, we examined the value of a DT model in the evaluation of adverse drug reactions, 14 an application that has been scarcely reported in the literature. We constructed a risk prediction model of vancomycin-associated nephrotoxicity and found that the predictive accuracy of DT and multiple logistic regression models (the conventional statistical method) were accurate (87.3% for both).…”

Changes of serum miR-223-3p in patients with oral cancer treated with TPF regimen and the prognosis

“…The diagnosis and treatment of OSCC are advancing due to the development in medical fields, but the prognosis is not very good. Data show that the 5-year survival rate of OSCC is only 50 to 60%, and some patients have advanced disease and local metastasis at the time of diagnosis, which is not conducive to clinical treatment (4,5). A previous study stated that the patients with advanced oral cancer had a poor 3-year survival rate (6).…”