2024 World Conference on Lung Cancer (WCLC) - ePosters-EP.04B.06 High-Resolution CT with 1024-Matrix for AI-Assisted Diagnosis System in the Evaluation of Pulmonary Nodules

EP.04B.06 High-Resolution CT with 1024-Matrix for AI-Assisted Diagnosis System in the Evaluation of Pulmonary Nodules

Pdf Summary

The study by Qinling Jiang focuses on improving the assessment and diagnosis of pulmonary nodules through high-resolution CT imaging using a 1024x1024 matrix in conjunction with an AI-assisted diagnostic system. Accurate evaluation of lung nodules is crucial for effective clinical management, and recent advancements in artificial intelligence (AI) have enhanced this process by increasing detection accuracy and reducing interpretation time for radiologists.<br /><br />This retrospective study analyzed CT chest scans of 344 patients from two hospitals between January 2020 and November 2023. Images were reconstructed using two different matrix resolutions, 512x512 and 1024x1024, to compare their efficacy in lung nodule assessment. The evaluation involved subjective image quality ratings from two radiologists using a 5-point Likert scale and the performance of an AI-assisted diagnostic system tasked with detecting various types of pulmonary nodules—namely solid, part-solid, and nonsolid nodules.<br /><br />The results revealed that the 1024x1024 matrix offered significantly higher image quality, with a subjective evaluation score of 4.18 compared to 3.63 for the 512x512 matrix. In terms of nodule detection accuracy, the 1024 matrix achieved superior results: 98.32% accuracy rate, 98.32% precision, and 100% recall, compared to 91.63%, 95.68%, and 95.59% respectively for the 512 matrix. The classification accuracy of various nodule types also improved markedly with the higher resolution matrix.<br /><br />In conclusion, adopting a 1024x1024 matrix for CT imaging significantly enhances image quality and detail, leading to more accurate AI evaluations of lung nodules. This improvement could contribute to better clinical decision-making and outcomes for patients.

Asset Subtitle

Qinling Jiang

Meta Tag

Speaker Qinling Jiang

Topic Screening and Early Detection

Keywords

pulmonary nodules

high-resolution CT

1024x1024 matrix

AI-assisted diagnostics

lung nodule assessment

image quality

nodule detection accuracy

retrospective study

clinical management

radiologist evaluation