Open AccessPenentuan Kurva Kelengkungan Tulang Belakang pada Citra X-ray Skoliosis Menggunakan Metode Fuzzy C-Means
Author(s) -
Bagus Adhi Kusuma
Publication year - 2019
Publication title -
jurnal media informatika budidarma/jurnal media informatika budidarma
Language(s) - English
Resource type - Journals
eISSN - 2614-5278
pISSN - 2548-8368
DOI - 10.30865/mib.v3i1.992
Subject(s) - cobb angle , scoliosis , segmentation , edge detection , artificial intelligence , enhanced data rates for gsm evolution , measure (data warehouse) , canny edge detector , fuzzy logic , computer science , cluster analysis , mathematics , image processing , computer vision , image (mathematics) , medicine , data mining , surgery
Scoliosis is a disorder that requires X-ray detection. Early detection is needed by patients with scoliosis. Based on information obtained through early detection, it will enable doctors to take the initial steps of treatment quickly. Determination of the spinal curve is the first step that serves to measure how severe the angle of the scoliosis curve. The angular severity of the scoliosis disorder can be calculated using the Cobb angle. Therefore, by estimating the spinal curve, we can also estimate the Cobb's angle. Based on the method from the previous study, the interobserver value can touch 11.8o with an error in the intraobserver measurement of 6o. Thus, during the Cobb Angle calculation process, subjective aspects are common and can still be tolerated today. But this aspect can also be the most frequent problem when using manual measurement methods by doctors. This study proposes an algorithm to measure the curve of the spine with computer-aided of X-ray images quickly with a error level error that is still within tolerance value. The data pre-processing process is carried out using Canny edge detection. The Fuzzy C-Means clustering algorithm (FCM) can detect the center point of the vertebral segment after segmentation of edge detection pre-processing. Formation of the spinal curve is done by the polynomial curve fitting method with the results of accuracy of 2.45o. Based on information on the spinal curve, the severity of the form of the scoliosis curve can be classified into four types, normal, mild, moderate and severe. Of the four levels, this system can be used to detect whether a person has scoliosis or not.