Improved Performance of Extraction Techniques on Boundary Based Centerline Algorithms for Identifying Appendicitis in Ultrasound (USG)

Digital Image Processing (DIP) performance continues to develop along with solving various problems in the health sector. This can be seen from the extraction technique used in the process of identifying appendicitis. Identifying this disease can be done by utilizing medical images such as Ultrasonography (USG) for early diagnosis. However, in this case, the medical image on ultrasound still has a lot of spike noise. Hence, the identification process requires the role of DIP in producing precise and accurate diagnostic results. This research aims to carry out the process of identifying appendicitis by improving the performance of the extraction technique using the boundary-based centerline algorithm (BBC). The improvement of the centerline algorithm in the extraction technique aims to represent the boundaries of an object's area with other objects. The research dataset uses 2097 ultrasound images from 90 patients sourced from the West Sumatra Lung Hospital. The performance results of the BBC algorithm in identification show a quite good improvement with an accuracy of 92%. These results can be based on the performance of the BBC algorithm which has been able to provide accuracy based on the role of boundary lines interactively in describing object identification. Overall, the improvement in extraction techniques carried out in this research contributes to DIP and becomes an alternative solution for medical parties in the process of diagnosing appendicitis.