Computer assisted diagnosis for diagnosis and monitoring of airway disease
The Erasmus Medical Center LungAnalaysis laboratory and artificial intelligence company Thirona jointly developed a fully automated method (LungQTM-AA) to measure dimensions of all visible airways and arteries on a lung computed tomography (CT) scan.
Airway abnormalities are an important feature of many lung diseases. Due to chronic infection and inflammation airway walls can thicken and widen. These airway abnormalities can be observed on a lung CT scan where a large number of airways and their accompanying arteries are visible. On a lung CT of a young child up to 150 airway artery pairs can be seen and in an adult up to 1000 pairs. For the sensitive detection and monitoring of airway abnormalities their dimensions should be measured and compared to the dimensions of the arteries. However, in clinical practice radiologists do not routinely measure the exact dimensions of airways and arteries as this is an extremely time consuming task which can take between one to five days for a single CT scan. For this reason LungQTM-AA was developed to fully automatically detect and measure dimensions of all airway-artery pairs on lung CTs of children and adults with cystic fibrosis (CF) which can show mild to very severe abnormalities.
LungQTM-AA was extensively tested in different groups of patients with CF. It was shown that LungQTM-AA is very sensitive to detect and monitor airway abnormalities both in children as in adults. Next, LungQTM-AA was also successfully tested in other lung diseases such as asthma and chronic bronchitis.
LungQTM-AA is already being used in clinical studies as outcome measure to see whether drugs have a positive effect on the progression of airway abnormalities. In addition, LungQTM-AA will be made available for patient care adding important objective quantitative information to detect airway disease and to evaluate the efficacy of treatment.
