METODE ASSOCIATION RULE UNTUK ANALISIS CITRA CT ORGAN PASIEN KANKER OVARIUM

METODE ASSOCIATION RULE UNTUK ANALISIS CITRA CT ORGAN PASIEN KANKER OVARIUM aDwina Kuswardani, bM.Rahmat Widyanto, c Indang Trihandini a,b Fakultas Ilmu Komputer, Universitas Indonesia,Depok 16424 c Fakultas Kesehatan Masyarakat, Universitas Indonesia,Depok 16424 E-Mail: a [email protected] Abstrak Penelitian ini melakukan analisis organ liver, ginjal dan uterus pasien kanker ovarium ada citra Computed Tomography (CT) scan. Analisis dari ketiga organ tersebut menggunakan Association Rule. Metode Association Rule merupakan bagian dari proses data mining yang bertujuan untuk menemukan kecenderungan suatu data, dalam hal ini dapat menemukan pola ciri organ liver, ginjal, dan uterus dari penderita kanker ovarium. Pembentukan Association Rule terdiri dari tiga tahap, yaitu ekstraksi fitur berdasarkan fitur ukuran dan intensitas organ, transaksi basis data, dan diakhiri penerapan Association Rule. Uji coba dilakukan terhadap tujuh belas data pasien kanker ovarium dari Rumah Sakit Kanker di Jakarta. Hasil dari Asociation Rule ditemukan bahwa ukuran uterus mempunyai pola ciri dengan support rule sebesar 55 %, pola ciri hubungan organ liver dan ukuran organ uterus dengan support rule 45 %, pola ciri hubungan organ liver, intensitas ginjal dan ukuran organ uterus dengan support rule 40 %, pola ciri hubungan organ intensitas ginjal dan ukuran uterus dengan support rule 50 %. Penerapan metode Association Rule untuk analisis citra CT scan organ pasien kanker ovarium ini diharapkan dapat membantu ahli medis dalam melakukan diagnosis. Kata Kunci: Association Rule, Computed Tomographi (CT), Citra Abdominal dan Pelvis. Abstract This research work on analysis of the organs liver, kidney anduterus, ovarian cancer patients at the image of Computed Tomography (CT) scan. Analysis of these organs use the Association Rule. Method of Association Rule is part of the process of data mining which aims to find the tendency of the data in this case can find a pattern characteristic of organs liver, kidney, and uterus of patients with ovarian cancer. Formation of Association Rule comprises three stages: feature extraction based on feature size and intensity of the organ, the transaction database, and terminated the application of Association Rule. Tests conducted on 17 patients with ovarian cancer data from cancer hospital in Jakarta. The result of Association Rule is found that uterus size has a characteristic pattern with support of rule 55 %, a pattern characteristic of relationship between liver organ and uterus size with support of rule 45 %, pattern characteristic of relationship between liver, kidney intensity and uterus size with support of rule 40 %, pattern characteristic of organ relationship kidney intensity and uterus size with support of rule 50 %. Application of Association Rule method for CT image analysis organ ovarian cancer patients is expected to help medical experts in making the diagnosis. Key words: Association Rule, Computed Tomographi (CT), Citra Abdominal and Pelvis