Novel Algorithm for the Classification of Emphysema Severity on CT Scans Download HRPP SOPs Clinical Trials Find a Researcher NJH ID: #19-03 Background Chronic obstructive pulmonary disease (COPD) is a progressive group of lung disease characterized by chronic inflammation of the lung that causes obstructed airflow, difficulty breathing, cough, mucus production and wheezing. If affects 12 million adults in the United States, with 12 million more thought to be undiagnosed. Emphysema and chronic bronchitis are the two most common conditions that contribute to COPD. Visual and quantitative computed tomography (CT) evaluation are currently the methods used for assessment of emphysema in COPD. The Fleischner Society proposed a structured system for visual classification that uses a six-point ordinal scale to grade parenchymal emphysema. Yet, this method is time consuming and subjective, requiring substantial training and being difficult to perform in routine practice. As a result, there is a need for a validated and faster automatic technique to more reliably identify and classify the severity of emphysema on CT scans. Technology Drs. Humphries and Lynch have developed and validated an objective computational method to identify and classify the severity of phenotypic abnormalities related to COPD. They used a deep learning algorithm for analysis and classification of emphysema on chest CT according to an established grading scale. This automatic analysis provides additional information compared to visual assessment and traditional quantitative CTs and could be used to predict the degree of impairment (pulmonary insufficiency) and mortality risk even in mild but clinically significant abnormalities that may be not apparent on visual assessment. More severe grades of emphysema severity have been associated risks of lung cancer, emphysema progression and overall mortality risk. Potential Applications This method provides the ability to objectively stratify the severity of emphysema on baseline CT and to evaluate for worsening on sequential CT scans. Potential clinical applications include lung cancer screening, where documentation of emphysema severity relates to cancer risk and may be a motivating factor for smoking cessation. In therapeutic trials objective classification of emphysema severity may contribute to participant selection and entry criteria, and may also be used to assess drug response. State of Development A functional prototype of the method has been applied to approximately 15,000 CT exams across multiple cohorts. Emphysema scores obtained from this method have consistently been correlated with physiologic impairment and mortality risk. Publications Humphries SM, et al. Deep Learning Enables Automatic Classification of Emphysema Pattern at CT. Radiology. 2020 Feb; 294 (2):434-444. PMID: 31793851 Lynch DA, et al. CT-based Visual Classification of Emphysema: Association with Mortality in the COPDGene Study. Radiology. 2018 Sep; 288(3):859-866 PMID: 29762095 Oh AS, et al. Visual emphysema at chest CT in GOLD stage 0 cigarette smokers predicts disease progression: Results from the COPDGene study. Radiology. 2020 Sep;296(3):641-9. PMID: 32633676 El Kaddouri B, et al. Fleischner Society Visual Emphysema CT Patterns Help Predict Progression of Emphysema in Current and Former Smokers: Results from the COPDGene Study. Radiology. 2021 Feb;298(2):441-9. PMID: 33320065 Oh AS, et al. Emphysema Progression at CT by Deep Learning Predicts Functional Impairment and Mortality: Results from the COPDGene Study. Radiology. 2022; 000:1–8 Radiology Patent Status Patent pending. Inventors Stephen Humphries, PhD and David Lynch, MD and Aleena Notary Licensing Status This technology is available for license. For Further Information, Contact: Emmanuel Hilaire, PhD Director Technology Transfer Office National Jewish Health 1400 Jackson Street, Room M206b Denver, CO 80206 Voice: 303.398.1262 HilaireE@njhealth.org