Evaluasi feature citra termografi kanker payudara dengan metode fraktal

• Main Author: ALAM, Wa Ode Siti Nur
• Format: Theses and Dissertations NonPeerReviewed
• Published: [Yogyakarta] : Universitas Gadjah Mada 2010
• Subjects: Health Information Systems (incl. Surveillance); Electrical and Electronic Engineering
• Online Access: https://repository.ugm.ac.id/85390/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=46254
• Institution: Universitas Gadjah Mada

Breast cancer is one kind of deadly diseases specially for women. Based on data of SIRS (Sistem Informasi Rumah Sakit) 2007, the percentage of breast cancer patients is 16,85% higher than serviks cancer which is 11,78%. Breast cancer become the first threat in America (suaramerdeka.com). Determinating breast cancer stage is very important for a doctor to determinate the best theraphy on a patient. The main aim of this research is to evaluate fractal feature for diagnosing breast cancer stages. This research used 12 sheets thermal images which known stages and with added noise. The processing of image steps are: creating grayscale images, contrast adjustment, cancer object cropping, black and white image converting, then counting the fractal dimension of the images. This research develops a Back Propagation neural networks to classify breast cancer stages base on the images fractal dimension. The result of this research shows that the fractal feature is very good for diagnosing/classifying breast cancer stages on thermography image. By using back propagation, it gave best classification accuracy of about 100% on image (with and without noise). The result occured for noisy image on the highest SNR.