Classification of Prostate Cancer with the Use of Artificial Immune System and ANN

• Published in: Genomics and Applied Biology Vol. 5; no. 3
• Main Author: Hasibe Cingilli, Vural
• Format: Journal Article
• Language: English
• Published: Richmond Sophia Publishing Group Inc 01.01.2014
• Subjects: Classification; Genetics; Immune system; Prostate cancer
• ISSN: 1925-1602; 1925-1602
• DOI: 10.5376/gab.2014.05.0003

Before analyzing cells in Laboratory in prostate cancer detection, a classification system can give valuable information about the cancer. The purpose of this paper is to assess the value of Artificial Immune System (AIS) and Artificial Neural Networks (ANN) for classification of prostate cancer cases. Paraffine-embedded prostate cancer tissue specimens of 50 prostate cancer subjects were used in this study. Age range was 35-72 years and all subjects were males. 10 subjects had family history of cancer and 40 patients were non family. An Artificial Immune System (AIS) which is based on clonal selection theory was used to classify these 50 subjects as healthy and patient. With the correct arrangement in system parameters, AIS has reached a classification accuracy of 93.33%. This ratio in 50 data means that in test phase, only one data was misclassified as healthy whereas indeed that data was belonging to a patient. The classification procedure was also done with another method which is a well-known effective classification method for biomedical data: Artificial Neural Networks. The result for this application was 100% with ANN method. While it seems that there is a big difference in the performances of AIS and ANN in the classification accuracy, this difference was only because of 1 data. Thus, it can be said that, AIS is also a good performing classification algorithm as well as ANN for this application.