Publications by authors named "Gustavo Antunes Rodrigues Duarte"
Article Synopsis
- The study focused on validating a deep learning model for bone age estimation in São Paulo, comparing it to the traditional Greulich and Pyle method.
- The model, based on a high-performing convolutional neural network, analyzed 714 hand and wrist radiographs, showing strong correlation with expert manual analysis (0.94) but indicating a mean absolute error of 7.68 months.
- While the algorithm was effective regardless of sex or race, it tended to overestimate bone age in younger patients, highlighting the need for further refinement.
