R. Selsi¹, Z. Rustam2, S. Gita3 and J. Pandelaki4

^{1, 2, 3} Department of Mathematics, FMIPA Universitas Indonesia,

Kampus UI Depok, 16424, Indonesia

(ruhul.selsi@ui.ac.id, rustam@ui.ac.id, sari.gita@sci.ui.ac.id)

 

⁴ Department of Radiology, RSCM Hospital, Jakarta, Indonesia

(jacubp@gmail.com)

ABSTRACT

 

Soft Tissue Tumor (STT) is currently being off the beaten track from tumor type of disease, or we called it rare to be found around the world. It is well known as an abnormal lump because of a new cell growth. There is no specific part of the body that can be affected but mostly it affects the area of the arms, hands, abdomen, and legs. Even this rare tumor can affect anybody in any kind of ages, commonly it affects the elder. According to some studies, the Soft Tissue Tumor is caused by the genetic mutation. The symptomps are different for each type of the tumor, depends on the type of the cells that have the genetic mutation. Later on, if someone is diagnosed with a Soft Tissue Tumor, the doctors will try to diagnose out if the disease has spread and how far the spreading is. This process is commonly called staging. In order to get the right treatment, the patient should detect and classify the stages with the doctors. Machine learning will help a lot in this process. The author proposes Fuzzy Robust Clustering method to classify the stages of Soft Tissue Tumor. One of the advantages of this method is : it does not require imputations. Even RFCM  (Robust Fuzzy C-Means) is even better than FCM (Fuzzy C-Means) because RFCM algorithms develops segmented result that more promising than the FCM algorithms. The result achieves up to 92,1 % accuracy so that this method can be a capable analysis tools.

 

Keywords: Classification, Robust Fuzzy C-Means, Soft Tissue Tumor