IAES International Journal of Artificial Intelligence (IJ-AI)
Vol 14, No 1: February 2025

Improved adaptive multi-threshold method for automatic identification of rhinosinusitis in paranasal sinus images

Putra, Ondra Eka (Unknown)
Sumijan, Sumijan (Unknown)
Tajuddin, Muhammad (Unknown)



Article Info

Publish Date
01 Feb 2025

Abstract

Rhinosinusitis, characterized by inflammation of the mucosa or mucous membrane within the paranasal sinuses, anatomical cavities situated in the facial bones, is the focus of this investigation. This study employs computed tomography (CT)-scan images comprising sagittal slices of the paranasal sinuses, acquired through a CT device featuring a Philips Ingenuity CT model MRC880 tube type, identified by tube serial number 163889, with a pixel value resolution of 0.24 mm. The primary objective of this research is to automatically identify and delineate rhizosinusitis-affected areas. This involves the application of multi-threshold values during the segmentation process, utilizing the improved adaptive multi-threshold (IAMT) segmentation method. The research dataset encompasses 380 slices of CTscans derived from 10 patients displaying indications of rhinosinusitis. Analysis of the test results reveals that the smallest observed rhinosinusitis size in this study is 0.05 cm2 on the right side, while the largest size measures 1.81 cm2 , yielding an accuracy rate of 96.66%. The magnitude of rhinosinusitis sizes serves as an indicative measure of the extent of inflammation within the paranasal sinus region, thereby suggesting a potential need for more intensive treatment interventions for the affected patients.

Copyrights © 2025






Journal Info

Abbrev

IJAI

Publisher

Subject

Computer Science & IT Engineering

Description

IAES International Journal of Artificial Intelligence (IJ-AI) publishes articles in the field of artificial intelligence (AI). The scope covers all artificial intelligence area and its application in the following topics: neural networks; fuzzy logic; simulated biological evolution algorithms (like ...