®®®® SIIA Público

Título del libro: Intelligent Data Sensing And Processing For Health And Well-Being Applications
Título del capítulo: User-centered data mining tool for survival-mortality classification of breast cancer in mexican-origin women

Autores UNAM:
GUILLERMO GILBERTO MOLERO CASTILLO; ISMAEL EVERARDO BARCENAS PATIÃ?O;
Autores externos:

Idioma:

Año de publicación:
2018
Palabras clave:

Classification; Data mining; Data mining in health applications; Logistic regression; Support vector machine; User-centered design


Resumen:

Nowadays, data mining is an important support technology for data analysis, the purpose of which is to discover information of interest in various fields of application, such as in health and well-being, which allows solving classification problems for forecasting trends and behaviors that occur in certain diseases. However, most current known data mining technologies are not oriented to user-centered design. Consequently, this results in data mining applications with limitations in usability and accessibility. In the health setting, this lack of user-centered data mining tools often results in misleading conclusions. In this work, it a classification system is proposed for user-centered analysis. We used this system to study the survival-mortality rate in Mexican-origin women diagnosed with breast cancer. The development of this proposed system is based on a methodology of user-centered data mining, which has as its foundation the principles of the ISO 9241:210:2010 standard. The system is composed of two classification algorithms, logistic regression and support vector machine. Furthermore, a confusion matrix is used to validate the outcome of the classification algorithms. We report tests on 2652 diagnosed cases in the Cancer Database of the Surveillance, Epidemiology, and End Results program of the National Cancer Institute. Results show 87% on average precision and 95% sensibility. As for the user-centered design, several oncological variables were identified by specialists of the General Hospital La Raza of Mexico City, among them laterality, behavior, grade, diagnostic confirmation, and RX Summ-Radiation. © 2018 Elsevier Inc. All rights reserved.


Entidades citadas de la UNAM: