Publications by authors named "Qi-Yuan Zeng"
Insights Imaging
May 2024
Article Synopsis
- An interactive, non-invasive AI system was developed to predict the malignancy risk associated with cystic renal lesions (CRLs) by utilizing a combination of 3D segmentation and classification models.
- The study analyzed data from 715 patients and used advanced techniques like a 3D-ResNet50 network and gated recurrent unit (GRU) for feature extraction and classification.
- Results showed outstanding performance in distinguishing benign from malignant CRLs, achieving high accuracy and sensitivity, thus enhancing clinical decision-making in diagnosing these lesions.
