Performance metrics achieved at the optimal threshold values (from Table 6), by the DenseNet-121 and VGG-16 models using calibrated (using the best performing calibration method from Table 3) and uncalibrated probabilities generated for Set-20, Set-60, and Set-100 datasets, constructed from the APTOS’19 fundus (n = 600) and Shenzhen TB CXR (n = 200) datasets, respectively.
收藏NIAID Data Ecosystem2026-03-13 收录
数据链接：
官方服务：
资源简介：
Performance metrics achieved at the optimal threshold values (from Table 6), by the DenseNet-121 and VGG-16 models using calibrated (using the best performing calibration method from Table 3) and uncalibrated probabilities generated for Set-20, Set-60, and Set-100 datasets, constructed from the APTOS’19 fundus (n = 600) and Shenzhen TB CXR (n = 200) datasets, respectively.
创建时间：
2022-01-27
