Diagnosis of ovarian cancer based on mass spectrum of blood samples

Hong Tang

Masters Thesis, Computer Science and Engineering Department, University of South Florida, 2003.

Abstract

The early detection of cancer is crucial for successful treatment, and medical researchers have investigated a number of early-diagnosis techniques. Recently, they have discovered that some cancers affect the concentration of certain molecules in the blood, which allows early diagnosis by analyzing the blood mass spectrum. Researchers have developed several techniques for the analysis of the mass-spectrum curve, and used them for the detection of prostate, ovarian, breast, bladder, pancreatic, kidney, liver and colon cancers.

We have continued this work and applied data mining to the diagnosis of ovarian cancer based on the mass-spectrum curve. We have identified the most informative points of this curve, and then used decision trees, support vector machines, and neural networks to determine the differences between the curves of cancer patients and healthy people.