A pioneering dual-scale AI model developed by researchers at Kaunas University of Technology (KTU) aims to enhance early lung cancer detection by analyzing CT scans more effectively. This system mimics a clinician’s approach, examining fine details and broader anatomical contexts concurrently. Trained on scans from healthy individuals and lung cancer patients, the AI can distinguish between normal tissue, benign changes, and malignant tumors, achieving over 96% accuracy. This dual approach is crucial as early detection significantly improves survival rates; five-year survival can increase from about 10% in advanced stages to over 90% in early detection. While the AI model shows promise, further validation in diverse, real-world settings is needed before clinical implementation. The technology is intended as a supportive tool for radiologists to enhance accuracy and reduce missed diagnoses, ultimately aiming to improve diagnostic workflows and patient outcomes in lung cancer screening.
For more insights, visit our Pinterest page!
Source link
