IMPLEMENTASI METODE SUM OF SQUARED DIFFERENCE DAN FUZZY C-MEANS UNTUK MENENTUKAN KURVA KEMIRINGAN TULANG BELAKANG SKOLIOSIS

Scoliosis is one of the disorders of the spine where the spine is bent sideways. One method for making sure someone determines scoliosis is to use Cobb's perspective on a spinal X-ray image. However, the Cobb angle measurement so far still requires a manual assist point. The importance of the spinal curve detection system in this study is to support the processing of Cobb's point of view so that it can be done automatically. This method is to estimate the position of the vertebrae of each subdivision. This research is intended for pre-processing in the system of determining the severity of computer-based scoliosis. Where the results of spinal curve testing are determined using the sum of squared difference (ssd) algorithm and fuzzy c-means. Furthermore, by using polynomial curve fittings of order 4, 5, and 6, the spinal curve can be estimated. This study discusses the performance of the SSD and FCM methods using parameters 8, 9, and 10 subdivision. The results of testing from 27 data of all spinal curves were successfully determined using SSD and FCM methods with the selection of 10- subdivision polymomial parameters of order 5 with 100% accuracy, and CWP values on SSD 37.65% with computation time of around 6 seconds, while CWP on FCM 33.78 % with a computing time of around 14 seconds. The analysis shows that using SSD or FCM algorithm on the parameter 10-subdivision polynomial order 5 scoliosis curves can be determined and can be used as pre-processing in Cobb angle measurement system.