Machine Learning Diagnostics System for Bronchiectasis

According to Australian data, more than 60% of adults with bronchiectasis have had symptoms since childhood. Early diagnosis is important to minimize later deterioration in lung function and provide timely care and treatment [1]. However, the amount of data collected by today’s HRCT scanners is beyond the ability of a radiologist to process in a normal clinical practice. To detect abnormal BA ratios in images, a system for computerized medical image processing, analysis and interpretation is required. Previously, some computer-aided diagnosis systems were developed to automate the analysis of bronchiectasis with promising results. Because the prevalence of bronchiectasis is age-related [1] [2], and geographic variations and quality of life conditions matters as well [1], a common automated system using the same criteria will not work everywhere. An adjustable system that is suitable for NT conditions will need to be developed.