PENDETEKSIAN MASSA BESAR PADA CITRA MAMMOGRAM POSITIF KANKER MENGGUNAKAN METODE REGION GROWING

Ghina, Shalihat (2015) PENDETEKSIAN MASSA BESAR PADA CITRA MAMMOGRAM POSITIF KANKER MENGGUNAKAN METODE REGION GROWING. Diploma thesis, UNIVERSITAS ANDALAS.

Text (Tugas Akhir Fulltext) 201501261107th_ta ghina shalihat - 1010953035 - teknik elektro.compressed.pdf Restricted to Repository staff only Download (2MB)

Abstract

Breast Cancer is one of the cancer that generally happened to women nowadays. The screening of breast cancer uses mammogram. Image from the result of mammogram screening is the image that used as detector of the condition and size of cancer mass. Mammogram is analyzed manually by the doctor or radiolog so it needs the education and also the experience about that study. It can makes the different interpretation among the doctor or radiolog in analyzing the mass of mammogram image. In this final project, it is explored the detection system of cancer mass where the system can detect the mass of mammogram image. This application can be used to support the opinion from doctor or radiolog in diagnostic the mass of mammogram image. The method that used in the process of mammogram image is region growing method. There are some steps in this system: converging of input image to grayscale image, reducing noise which used median filter, separating the object with the background, and uniting the area which suspected as cancer mass using region growing. Input image that used in this system is about 55 data of positive cancer and 15 data of normal condition. The output from this system is the image that shows the part of cancer mass in the input image which used here. The system evaluation comes from the analysis of paramedic using 11 output image consists of 10 positive cancer image and 1 normal image. The percentage of compatibility from 11 output system which paramedic analyzed is 90,9 %. Then, it is obtained the test with 40 mammogram image to get parameter which consists of the pixel sum in output of object diameter that states whether it is the part of cancer or not. The part of output which stated as cancer has the pixel sum around ≥ 62 pixel of object diameter and ≤ 57 pixel sum which stated as not a cancer. This parameter is used to evaluate another 30 mammogram image. It uses the determine parameter, where it explains whether it is the part of the cancer or not part of the cancer in result of the output is qualified or not qualified in the parameter. The percentage of compatibility in output system using the parameter is 93,3%, where 28 of 30 additional image fulfill the determine parameter.

Actions (login required)

View Item