Chung-Feng Jeffrey Kuo1, Yi-Shing Leu2, Richard Kuo3, Chin-Hui Su2, Tzu-Chieh Yuan4, Bo-Han Ke4, Nain-Ying Wu5. 1. Graduate Institute of Automation and Control, National Taiwan University of Science and Technology, Taipei 10607, Taiwan. Electronic address: jeffreykuo@mail.ntust.edu.tw. 2. Department of Otolaryngology of MacKay Memorial Hospital, Taipei, Taiwan. 3. Department of Radiology in Mackay Memorial Hospital, Taipei, Taiwan. 4. Graduate Institute of Automation and Control, National Taiwan University of Science and Technology, Taipei 10607, Taiwan. 5. Department of Materials Science and Engineering, National Taiwan University of Science and Technology, Taipei 10607, Taiwan.
Abstract
BACKGROUND AND OBJECTIVE: In this study, we aim to develop a system that uses computed tomography (CT) imaging for three-dimensional (3D) reconstruction of the trachea as therapy for tracheal stenosis in infants, and further calculate the cross-sectional area and volume, assisting doctors in clinical diagnosis. METHODS: We first used image processing, calculating the cross-sectional area and volume. We used the improved median filter for image processing and designed the system for capturing the cross-sectional area of endotracheal tube. We then established 3D reconstruction images with isosurface extraction technology and calculated the cross-sectional area and volume. Medical indicator data analysis was performed. RESULTS: The median filter developed in this study performed better in filtering speckle noise compared to traditional filtering methods. Furthermore, the median filter can keep fine texture feature, so that the subsequent contour selection and 3D reconstructed volume are more accurate. We also proposed new medical grading indexes according to tracheal obstruction volume ratio to assist doctors with the diagnosis and provide recommendations on treatment. Seventeen samples were examined in this study. Four sections of each sample are reviewed. Sixty-eight sections were used for validation, and the overall accuracy is very reliable. CONCLUSIONS: Using image processing we obtained tracheal CT images before 3D reconstruction and calculated the cross-sectional area and volume of the trachea. New medical indicators are proposed according to the location and severity of stenosis to assist doctors with diagnosis.
BACKGROUND AND OBJECTIVE: In this study, we aim to develop a system that uses computed tomography (CT) imaging for three-dimensional (3D) reconstruction of the trachea as therapy for tracheal stenosis in infants, and further calculate the cross-sectional area and volume, assisting doctors in clinical diagnosis. METHODS: We first used image processing, calculating the cross-sectional area and volume. We used the improved median filter for image processing and designed the system for capturing the cross-sectional area of endotracheal tube. We then established 3D reconstruction images with isosurface extraction technology and calculated the cross-sectional area and volume. Medical indicator data analysis was performed. RESULTS: The median filter developed in this study performed better in filtering speckle noise compared to traditional filtering methods. Furthermore, the median filter can keep fine texture feature, so that the subsequent contour selection and 3D reconstructed volume are more accurate. We also proposed new medical grading indexes according to tracheal obstruction volume ratio to assist doctors with the diagnosis and provide recommendations on treatment. Seventeen samples were examined in this study. Four sections of each sample are reviewed. Sixty-eight sections were used for validation, and the overall accuracy is very reliable. CONCLUSIONS: Using image processing we obtained tracheal CT images before 3D reconstruction and calculated the cross-sectional area and volume of the trachea. New medical indicators are proposed according to the location and severity of stenosis to assist doctors with diagnosis.