Network-Based AI Engine Performs Airway Segmentation from CT Images

A new 2.5D convolutional neural network (CNN)-based artificial intelligence (AI) engine enables accurate airway segmentation from computed tomography (CT) images without any human interaction.

Developed by Coreline Soft Co. Ltd. (Seoul, Rep. of Korea), the COPD analysis solution named AVIEW Metric-Lung offers AI-powered lung/lobe segmentation, airway measurement and INSP/EXP lung registration for various quantifications. Designed for quantitative image biomarker of COPD, the software uses chest CT images to provide various quantitative analysis reports.


It provides seven cutting-edge methods to analyze the conditions of lung parenchyma, airway, and lung vessels that affect the lung function of the patient: LAA, size-based LAA, airway characteristic, INS-EXP parametric map, air-trapping, lung vessel distribution and ILD classification analysis methods. Easy workflow minimizes user interaction during lung/lobe segmentation, airway segmentation and, elastic inspiration/expiration registration.

Since thousands of quantitative results make it difficult to interpret the lung function of the patient, AVIEW Metric-Lung provides intuitive charts and visualizes groups of values to the lung anatomical structure for a comprehensive understanding. It helps to reduce the inconsistencies between readers and accurately determine how far the disease has progressed.

Using fast, high-quality 3D rendering, AVIEW Metric-Lung figures out more than 1,400 quantifications per case that can all be exported in CSV format for further research. It performs all quantifications without a single click, making it the first of its kind solution in the world. The web-based software can be accessed anywhere with any device using a web browser.

Related Links:
Coreline Soft