Klasifikasi Penyakit Polyp Kolon Melalui Citra Gastrointestinal menggunakan Deep Belief Network

Abstract

Polyp is a disease in the form of protrusion on a surface on the mucosa. Colonic polyps protrusions abnormal that occur in the mucosa towards the lumen, which generally have a stalk or sessile. The trigger for colon polyps is due to unhealthy diet and lifestyle, smoking is also a source of colon polyps. In classifying colonic polyps, a biopsy by taking a sample of cells or tissue for analysis in the laboratory. Examination is still done manually by placing it on the preparation and looking at it through a microscope, then comparing it to observing cells, in this case it will take inefficient. Therefore, a classification disease colonic polyp using the method Deep Belief Network is the goal of this research. The data used in this study were 900 data. In this study, the dataset will pre-process and classify the dataset using the Deep Belief Network method. It can be concluded that the Deep Belief Network method used using the ability to classify colonic polyps in the form of stem polyps, flat polyps and non-polyps with gastrointestinal images has an accuracy value of 93% on the system.