2022 World Conference on Lung Cancer (ePosters)-EP01.03-003. AI Detection of Emphysema in an Ultra-Low Dose CT Lung Cancer Screening Program

EP01.03-003. AI Detection of Emphysema in an Ultra-Low Dose CT Lung Cancer Screening Program

Pdf Summary

This study explored the use of Artificial Intelligence (AI) software in detecting emphysema as part of a lung cancer screening program. The researchers compared the effectiveness of AI and radiologist assessment in diagnosing emphysema. They used CT scans and spirometry to evaluate the accuracy of AI and radiologist scoring. The study found that AI was more sensitive in detecting possible emphysematous anomalies but correlated less strongly to spirometrically defined chronic obstructive pulmonary disease (COPD) compared to radiologist scoring. The presence of fibrosis negatively affected both AI and radiologist scoring, but AI was impacted more significantly. The researchers concluded that AI software can be used to detect emphysema using ultra-low dose CT scans, but its reliability compared to standard radiologist assessment needs to be further established. The study included CT scans of individuals with radiologist-defined emphysema and non-emphysema participants for comparison. The AI scores for emphysema were more frequent and higher compared to radiologists, but radiologists had better distinguishing capacity and better correlation with spirometry and DLCO measures. The presence of fibrosis diminished the ability of both AI and radiologists to differentiate cases with decreased DLCO, but this effect was more pronounced in AI. Overall, the study suggests that AI can be a useful tool in detecting emphysema in a lung cancer screening program, but further research is needed to fully understand its diagnostic effectiveness.

Asset Subtitle

Chellan Kumarasamy

Meta Tag

Speaker Chellan Kumarasamy

Topic Early Detection and Screening - Implementation

Keywords

Artificial Intelligence

emphysema detection

lung cancer screening

AI software

radiologist assessment

CT scans

spirometry

chronic obstructive pulmonary disease

fibrosis

ultra-low dose CT scans