Identifying High-Risk Breast Cancer Patients Using Microarray And Clinical Data / Nur Azni Nasuha Bt Ngisa

Abstract in English

"A report project submitted in partial fulfillment of the requirements for the award of Bachelor of Computer Science (Hons)." -- On t. p.

Project paper (Bachelor of Computer Science ) - University Malaysia of Computer Science and Engineering, 2017.

Recently, most of the researchers start to use microarray technology to screen the whole gene expression for breast cancer prediction. The advent of this technology into breast cancer prediction has shown significant impact on the performance and optimization on the data used. Previous studies have demonstrate the potential value of gene expression in accessing the risk and used to identify the subgroups of breast cancer patients either high- risk or low-risk. Due to that, the application of clinical data appears to be underused in cancer prediction. Up to the present, the study on a complete molecular classification of breast tumours is still ongoing for further clarification due to the existence of the curse dimensionality in microarray data. Therefore, this study proposed develop a prediction method, GridPCA integration to identify high-risk breast cancer patients using microarray and clinical data with the aim to have a more accurate result in prediction compare to the previous works. Several additional steps will be added into the proposed method to deal with the high dimensionality of microarray data. Besides, a more practical strategy will be identified to utilize both, clinical and genetic markers in order to improve the prediction accuracy. By going through a few processing steps after being integrated, the average predictive accuracy for GridPCA integration achieved approximately 82% with Decision Tree classifier. In short, the result obtained could help the doctors in making decision for tailoring appropriate treatment to their patients and increasing the survival rate of breast cancer patients.