Deep-Learning Algorithm for CT Predicts Functional Impairment and Mortality from Emphysema

Patients with an increasing emphysema severity as graded by a deep learning algorithm on sequential CT scans in cigarette smokers were found to have an increased functional impairment and increased risk of mortality.

The analysis of the prospective Genetic Epidemiology of Chronic Obstructive Pulmonary Disease (COPDGene) study participants published in Radiology included baseline and five-year follow-up CT scans from 2007 to 2017. The deep learning algorithm, based on convolutional neural network and long short-term memory architectures, was trained to classify patterns of parenchymal emphysema according to Fleischner Society scoring system.

Of the 5,056 participants, 26% had emphysema progression at five-year follow-up with progressive airflow obstruction, a greater decline in 6-minute walk distance and greater progression in quantitative emphysema extent than those who did not progress. A higher mortality rate was also found in the group with emphysema progression, based on a multivariable Cox regression analysis.

All participant underwent a volumetric, non-contrast axial inspiratory and expiratory CT using a standardized protocol. A dedicated software program, LungQ, was used for quantitative analysis of emphysema extent. A deep learning algorithm developed and validated using  Python was trained on categorical visual scores assigned by two trained research analysts, with disagreements greater than one grade adjudicated by a thoracic radiologist with more than 30 years of experience.

The authors suggest their findings extend  prior study results that have established the significance of baseline parenchymal emphysema pattern according to the Fleischner Society scoring system using visual methods, and its role in predicting future emphysema progression and mortality. They wrote that an increase in emphysema severity score at five years was an independent predictor of disease progression and mortality and the results suggest the clinical value of automatic, structured grading of emphysema severity at CT for identification of patients at greater risk.