2023 World Conference on Lung Cancer (Posters)-P1.27. Artificial Intelligence Analysis Predicts Highly Malignant Potential in Small-Sized Lung Adenocarcinoma

P1.27. Artificial Intelligence Analysis Predicts Highly Malignant Potential in Small-Sized Lung Adenocarcinoma - PDF(Slides)

Back to course

Pdf Summary

A study conducted at Tokyo Medical University analyzed the use of artificial intelligence (AI) technology to evaluate small-sized lung adenocarcinomas. The study aimed to objectively assess radiological findings using three-dimensional (3D) imaging analysis with AI software. A total of 207 patients with clinical stage N0 non-small cell lung cancer were included in the study.<br /><br />The results showed that tumors classified as "AI-solid" by the AI software had a high incidence of lymph node metastasis and were pathologically upstaged in comparison to non-AI solid tumors. Specifically, 19% of AI-solid tumors had lymph node metastasis, while only 2% of non-AI solid tumors exhibited this characteristic. Additionally, 49% of AI-solid tumors were pathologically upstaged, compared to only 4% of non-AI solid tumors.<br /><br />The study also identified the association between radiological findings analyzed by the AI software and pathological factors. The median consolidation-tumor ratio was 0.67, and the median pathological invasive size of the tumors was 0.7 cm. The study included patients with a range of adenocarcinoma subtypes, with the majority being AIS/MIA or invasive adenocarcinoma.<br /><br />The findings suggest that the use of 3D imaging analysis with AI technology enables an objective evaluation of radiological findings in small-sized lung adenocarcinoma. It highlights the importance of carefully considering the indications for sublobar resection, as tumors classified as AI-solid have a higher likelihood of lymph node metastasis.<br /><br />Overall, this study demonstrates the potential of AI analysis in predicting the malignancy levels of small-sized lung adenocarcinomas, leading to more accurate and informed treatment decisions.

Asset Subtitle

Yujin Kudo

Meta Tag

Speaker Yujin Kudo

Topic Early-Stage NSCLC: Innovation & New Technology

Keywords

Tokyo Medical University

artificial intelligence

AI technology

lung adenocarcinoma

3D imaging analysis

lymph node metastasis

pathological upstaging

radiological findings

sublobar resection

malignancy prediction