2026 IASLC CT Screening and Early Detection Symposium-WS02.15 - Talk 1: Quantification and Metrics in Thoracic CTs

Pdf Summary

The document discusses the rapid growth of quantitative and AI-based metrics in thoracic CT and emphasizes that measurements are only valuable if they are accurate, reproducible, clinically useful, and quality-controlled across software versions and patient populations.<br /><br />Key screening-related CT metrics include:<br />1. Abnormal actionable findings with evidence-based guidelines,<br />2. Abnormal but uncertain findings,<br />3. Non-actionable abnormalities,<br />4. Body composition measures,<br />5. Predictive measurements.<br /><br />A major theme is prediction of future disease from CT images. The Sybil deep learning model uses a single low-dose CT scan to predict lung cancer risk up to 6 years later, with strong performance in multiple datasets and validation cohorts. Other predictive approaches use radiomics, reinforcement learning, and generative AI to classify nodule risk or simulate how nodules may evolve over time.<br /><br />The presentation also highlights opportunistic coronary artery calcium (CAC) detection on non-gated chest CT. AI can rapidly quantify CAC, enabling better cardiovascular prevention. The NOTIFY-1 study showed that notifying clinicians about incidental CAC substantially increased statin prescribing compared with usual care.<br /><br />For interstitial lung disease, the document distinguishes fibrotic ILD from incidental interstitial lung abnormalities (ILA), which may progress over time and warrant pulmonary evaluation and follow-up. AI tools such as IQ-UIP can detect usual interstitial pneumonia with good sensitivity and specificity, helping identify high-risk patients.<br /><br />Body composition metrics are another focus, especially bone mineral density. Opportunistic CT-based screening for osteoporosis can detect low BMD and compression fractures, complementing underused DEXA screening. Studies showed that CT-based BMD measurement can work across scanner types and protocols, and large Medicare analyses suggest substantial increases in screening rates and potential cost savings if opportunistic CT screening were implemented.<br /><br />Overall, the document argues that thoracic CT offers many opportunities for actionable quantitative metrics, especially when linked to clinical pathways and preventive care.

Asset Subtitle

Ella Kazerooni

Meta Tag

Concept Thoracic CT

Concept Quantitative Imaging

Concept Lung Cancer Risk

Concept Coronary Artery Calcium Score

Concept Interstitial Lung Disease

Speaker Ella Kazerooni

Keywords

thoracic CT

quantitative metrics

AI-based analysis

lung cancer prediction

Sybil model

coronary artery calcium

interstitial lung disease

opportunistic screening

bone mineral density

predictive radiomics

Thoracic CT

Quantitative Imaging

Lung Cancer Risk

Coronary Artery Calcium Score

Interstitial Lung Disease

Ella Kazerooni