Test data and models to the paper "Assessment of glomerular morphological patterns by deep learning algorithms". Different, from other groups, defined CNN-models (saved as .pt-files) are trained to identify nine predefined patterns of glomerular changes. The models are: AlexNet [1], ResNet18-152 [2], ResNet34 [2], ResNet50 [2], ResNet101 [2], ResNet152 [2], vgg11 [3], vgg16 [3], vgg19 [3], squeeznet [4], inception [5], and densenet121 [6]. The patterns are pattern 01: normal glomerulus, pattern 02: amyloidosis, pattern 03: nodular sclerosis, pattern 04: global sclerosis, pattern 05: mesangial expansion, pattern 06: membranoproliferative glomerulonephritis (MPGN), pattern 07: necrosis, pattern 08: segmental sclerosis, and pattern 09: other structures / default.

References:

Krizhevsky, A., One weird trick for parallelizing convolutional neural networks. arXiv preprint arXiv:1404.5997, 2014. He, K., et al. Deep residual learning for image recognition. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2016. Simonyan, K. and A. Zisserman, Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014. Landola, F.N., et al., SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size. arXiv preprint arXiv:1602.07360, 2016. Szegedy, C., et al. Rethinking the inception architecture for computer vision. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2016. Huang, G., et al. Densely connected convolutional networks. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2017.