2023 World Conference on Lung Cancer (Posters)-EP04.04. Relevance of AI (qCT ) In Automated Progression Monitoring of Pulmonary Nodules in Follow-up Lung CT Scans

EP04.04. Relevance of AI (qCT ) In Automated Progression Monitoring of Pulmonary Nodules in Follow-up Lung CT Scans - PDF(Abstract)

Back to course

Pdf Summary

This study explores the use of artificial intelligence (AI) in automating the tracking and monitoring of pulmonary nodules on low-dose computed tomography (LDCT) scans. The researchers utilized a dataset from the National Lung Screening Trial and employed deep learning algorithms for nodule detection and segmentation. They also developed an image registration method to align multiple CT screenings and track the nodules across time.<br /><br />The results show that the AI algorithm successfully detected and tracked a significant majority of the nodules in follow-up screenings. Out of 139 nodules identified in subsequent screenings, the algorithm correctly tracked 98 of them. Additionally, it accurately tracked 26 out of 35 nodules in the first follow-up. It was observed that around 32.6% of the tracked nodules exhibited progressive enlargement in volume, indicating potential malignancy. Furthermore, about 20.4% of the nodules showed shrinkage in volume in subsequent screenings.<br /><br />The study concludes that AI can effectively automate nodule tracking with high accuracy on LDCT scans. This technology can assist radiologists in identifying and monitoring the progression of pulmonary nodules that may become malignant in subsequent follow-ups. The use of AI algorithms can enhance the performance and efficiency of physicians in managing lung nodule-positive patients. It also has implications for improving patient management in lung cancer diagnosis and potentially reducing lung cancer-related mortality and morbidity.<br /><br />In summary, this research demonstrates the relevance of using AI, specifically qCT, in automated progression monitoring of pulmonary nodules. The findings highlight the potential of AI to aid in early detection and management of lung cancer, benefiting both healthcare professionals and patients.

Asset Subtitle

Rohitashva Agrawal

Meta Tag

Speaker Rohitashva Agrawal

Topic Screening & Early Detection: Nodule Management

Keywords

artificial intelligence

pulmonary nodules

LDCT scans

deep learning algorithms

nodule detection

image registration method

follow-up screenings

malignancy

patient management

lung cancer diagnosis